is The Effects
hp dh of Continued
Care

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

AE 0 A IO ER RR MERE RA BRB y
fo fg 1 i
BIER EIN ERIE EE EEE EEE n
1 Lr Ye rr mT TCT ITI

3
0 ETRE A 1 ln same
( IIII JRNENNIDEEOmIMIEmEEEE Ei mn ®
RUSE SEI08a0080 ERAN 00 NEE on
IIIT I SR ]
40 a ]
TIT map Tr HiNEMDERER |
0H 0 0 50 1 ke a )
MCEEN BIH TRANON ERI SE
(ITT A SL 0 AO OE 1 (V ;
10 0 0 1 A A RR TR UE A SS a es
iL BR SE n or A
E1041. 0 OS SR 0 5 1 8 1, 10
1aasnIIINaREIRNmERIEAnnInaRIRRRRInNEN Si A RS Bm IY ae
#40 : ows Ni’
I OL 50 1 OR rN
OO i 1 1 ene IE ERE REE
DONEREEAIERIENIARER nA mEmERRLE TT [ERR Em 1
[I' "Try NM hpuEtiiaEmeS MIB EaIe

- {

UNITED STATES DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
LIS Health Services and Mental Health Administration od
rial Center for Health Services Research and Development

 

 
 

 

TT I ——

 

 
EFFECTS OF CONTINUED CARE:

N'\

A Study of Chronic lllness in the Home

Sidney Katz, M.D., Professor and Director, Office of Health
Services Education and Research, College of Human
Medicine, Michigan State University

Amasa B. Ford, M.D., Associate Professor of Community Health,
School of Medicine, Case Western Reserve University

Thomas D. Downs, Ph.D., Associate Professor of Biometry,
School of Public Health, University of Texas

Mary Adams, Ph.D., Associate Professor of Public Health Nursing,
Frances Payne Bolton School of Nursing,

Case Western Reserve University

Dorothy I. Rusby, R.N., Director, Visiting Nurse Association

of Cleveland pd

Contract No. HSM110-69-240

DHEW Publication No. (HSM) 73-3010

December 1972

DEPARTMENT OF HEALTH, EDUCATION, AND WELFARE
up Health Services and Mental Health Administration_|
Natio

nal Center for Health Services Research and Development
CONTENTS

Preface . . . . . . eee

PART
The Plan, Elements, and Results of the Study

Chapter I. Introduction . . . . . . . ....
Chronic Disease in American Society . . . . . . . .. .....
Home Health Services After Rehabilitation . . . . . . . . . ..
The Research Approach and Research Questions . . . . . . . .

Chapter II. The Research Plan . . . . . . . . .............
Research Design. . . . . . . . . . ...............
Study Intake . . . . . . LLL
Interval Observations . . . . . . . . . . ............
Deaths . . . . . . .
Terminal Observations . . . . . . . . . . . ...........
Organization of the Staff . . . . . . . . ............
Data Collection . . . . . . . . . . . . . ............

Chapter III. The Patient Sample . . . . . . . . . ...........
The Hospital . . . . . . . 0...
Sample Selection . . . . . . LLL.
2
SEX . oe ee eee
Disease . . . . . . LLL eee eee
Physical Disability . . . . . . .. ................
Socioeconomic Status and Function . . . . . . . ........
Social Interaction . . . . . . .. LL.
Social Deprivation . . . . . . LLL.
Psychological Characteristics . . . . . . . ee ee

Chapter IV. Providing Continued Care . . . . . . . . ........
The Visiting Nurse Association (VNA) . . . . . . .. ......
Service by Visiting Nurses in the Research Program . . . . . . .
Extent of Nurse Follow-Ups (Examples) . . . . . ........

iii
Effects of Continued Care

Chapter V. Interviewing and Measures . . . . . . . . .........
Interviewing . . . . . . . Lo...
Frequency and Completeness of Interviews . . . . . . .....
Measures . . . . . o.oo ee ee eee ee eee
Measures of Physical Function and Injury,

and Measuresat Death . . . . . . . . . .... ........
Social and Economic Measures . . . . . . ............
Psychological Measures . . . . . . .. .............
Measures of Care . . . . . . . . . . .

Chapter VI. Classification of Disease, Disability and Mental Status
atIntake . . . . . LLL
Classes Based on Principal Clinical Diagnosis . . . . . . . .. ..
Classes According to Disease-Disability Profiles . . . . . .. ..
Classes Based on Indicators of Major Chronic Conditions

Chapter VII. Effects of Treatment Program: All Patients and
Subgroups by Disease, Disability and Mental Status
Areas of Measurement . . . . ... LL...
Outcome in Terms of Function, Injury, and Mortality . . . . . .
Outcome in Terms of Use of Services . . . .. . ........

Chapter VIII. Classification of Social Factors for the Study . . . . . .
Identifying Social Characteristics at Intake . . . . . . . .. ..
Social Interaction at Intake . . . . .. .............
Social Deprivation at Intake . . . . ...............

Chapter IX. Effects of Treatment Program: Subgroups by Age,
Sex, and Social Characteristics . . . . . . ........
Outcome in Terms of Function, Injury, and Mortality . . . . . .
Outcome in Terms of Use of Services . . . . . .........

Chapter X. Interval Results . . . . . .................
Unbiased Interval Results . . . . ................

Chapter XI. Interpretations . . . ...................
Observation and Interaction Effects . . . . . . . ........
Effects of the Nursing Program . . . . . . .. .........

- PART II
Methods

Chapter XII. Statistical Methods . . . ................
The Experimental Design . . . . ................
Experimental Effects . . . ...................

Page

46
46
47
49

50
50
51
52

54
56
56
57
Page

Intake Classes and Effectiveness of Randomization . . . . . . . 98

Validity of the Experimental Groups . . . . . . . .. . .... 100

Outcome Classes . . . . . . . . oi iii 102
Chapter XIII. Evaluation Schedule, Study Forms and Supplementary

Definitions . . . . . . . . . LL... 106

Chapter XIV. Intake and Outcome Variables . . . . . .... .... 153

APPENDIX

Statistically Significant Treatment Effects at End of Study . . . . . . 164

Contents
 

 

 
PREFACE

In September 1961, a group of doctors and nurses met at Abington House (formerly
the Benjamin Rose Hospital) in Cleveland to plan a controlled study of the effects of
public health nursing on the welfare of patients discharged from a chronic disease
rehabilitation hospital. Three currents of interest converged at that meeting, each of
which contributed to the purpose and the final form of the study. Physicians,
nurses, and biometricians formed a steering committee which offered different
professional viewpoints and personal experiences. Together, the group reviewed the
relatively small number of related studies which had been published to that date.
From this background, the experiment reported in this book was developed, namely,
a controlled study of the outcome of a sustained program of home nursing service
for patients discharged from a chronic disease rehabilitation hospital.

For six years prior to the present project, Abington House had been the setting
for a series of ‘“‘Multidisciplinary Studies of Illness in Aged Persons”. Faculty
members of Case Western Reserve University School of Medicine started these
studies in 1954 with the recognition that objective and tested measures of function
which could be used in the chronological study of chronic illness, though
increasingly needed, were almost nonexistent. A sequence of continuing long-term
observations on patients with strokes and hip fractures, later expanded to include
patients in nursing homes and in a home care program, was undertaken in order to
develop and test such measures. Many specific observations, contributed by
members of several health professions involved in caring for these patients, were
systematically sifted and correlated in the search for reliable and sensitive measures
of function. Two useful indices, one of Independence in Activities of Daily Living
(JAMA. 85:914-919, 1963) and one of Independence in Socioeconomic Func-
tioning (J. Chron. Dis. 13:453-464, 1961), had been developed and were available
for use in the present study. Since the ‘“‘Multidisciplinary Studies” had been
conducted primarily on patients in and discharged from Abington House, a great
deal of information was available about this particular population. Methods for the
coilection of information about these patients were well developed and in current
use, and a staff of experienced research assistants was available.

Abington House, a unit of University Hospitals of Cleveland, is a small hospital
that provides rehabilitation and hospital services for patients with chronic illness.
The care of older persons is emphasized. Staff of the hospital had a particular
interest in the care of chronically ill patients after their discharge from the hospital.
Just prior to the present study, efforts had been made to bring discharged patients
back to a clinic at the hospital for follow-up evaluation. This ambulatory program
proved feasible for only half the patients. The other half could not return because of
physical disability and transportation problems or would not because they or their

vii
Effects of Continued Care

physicians did not consider that such reevaluation was needed. Many of the patients
who could not return to the hospital were those who were most ill or disabled, and
there appeared to be a need for extending continued services to these persons in
their homes, particularly with regard to nursing services. The hospital staff had been
accustomed to refer about 14 per cent of the patients to the Visiting Nurse
Association at the time of discharge. It was, therefore, considered logical to extend
this practice to meet the apparent needs of more of the discharged patients. The
research interest and experience of the staff made it possible to conduct this
expansion of services under controlled, experimental conditions which would permit
the objective measurement of the effect of these services on the welfare of the
patients to whom they were offered.

Members of the Department of Biometry of Case Western Reserve University
School of Medicine were actively involved in the study from the beginning. They
were challenged by the task of designing a controlled study of a health services
program and by its special problems of data handling and analysis. They were
responsible for quality control over data and for the appropriateness of inferences
drawn on the basis of the study’s design.

Another prime motive for the study came from the nursing profession. The
faculty of the Frances Payne Bolton School of Nursing recognized that the care of
the chronically ill will be an increasingly important part of nursing in the immediate
future. They wished to contribute both to nursing education and to better
understanding of the sociology of the elderly and chronically ill. In addition to
direct collaboration by the nurse-sociologist of the research team, faculty members
of the School of Nursing served on the study’s steering committee and as advisors
and consultants. A major contribution from the nursing profession was the
involvement of the Visiting Nurse Association of Cleveland. Staff members of this
community agency wanted to know which types of patients in the growing class of
the “chronically ill” can profit most by the services of the public health nurse. They
expressed the need for more information about the needs of patients after discharge
from the hospital and the hope that the study could go beyond the testing of
whether or not public health nursing has a measurable effect on the patient’s
functional level to the correlation of specific aspects of nursing with patient
progress.

The study, thus, had its origin in theoretical and practical questions posed by
physicians, nurses, and biometricians. The particular medical focus originated in
teaching and research experiences with hospital and home care of chronically ill
patients.

Acknowledgements

From the start of the study to its completion almost ten years later, this project was
made possible by many people. As a prelude to the book, we wish to express our
indebtedness to those who were not in a position to be recognized as authors and to
whom so much credit is due. Foremost among them were the three hundred patients
who must remain unnamed and who permitted intrusion into a bit of their lives in
the hope of contributing to the welfare of others.

Throughout the study, a series of advisors and consultants contributed time and
knowledge needed for the project. In the planning stages, Dr. George F. Badger,
Professor of Biostatistics, and Dr. Alexander L. Cicchinelli, Assistant Professor of
Biostatistics, helped formulate the research design. As the study progressed, Dr.
Badger made available his Department’s key punching, electronic computer, and
other data processing resources. Dr. Donald Mainland, Professor of Medical Statistics
at New York University School of Medicine, also reviewed the experiment design
critically. Dr. Harry Posman, Project Director of a related study for the Community
Service Society of New York, lent valuable assistance by extending his research
experiences during the planning and design stages. Other experts were asked to help
establish instruments and procedures in selected areas. Dr. August B. Hollingshead,
Professor of Sociology, Yale University, served as advisor in matters relating to the
measurement of social class; and Dr. Stephen L. Fink, Associate Professor of
Organizational Sciences, Case Institute of Technology, was the project’s consultant
in psychology. Representing the nursing profession as advisors were Margene O.
Faddis, Professor of Nursing, and Dr. Jean L. Stair, Professor of Public Health
Nursing, both from the Frances Payne Bolton School of Nursing. Dr. Stair served on
the project’s steering committee throughout the study, contributing much wisdom
and support. Dr. John H. Dingle, Professor of Preventive Medicine at Case Western
Reserve University School of Medicine, and Dr. Faye G. Abdellah of the U.S. Public
Health Service gave valuable advice during the progress of the study and throughout
the stages of manuscript development. Others in the University who encouraged and
advised the investigators on a continuing basis were Rozella M. Schlotfeldt, Dean of
the Frances Payne Bolton School of Nursing, and Dr. Robert H. Ebert, Professor of
Medicine.

Direct interviewing and evaluation was the responsibility of a team of research
nurses, physicians, and sociologists, who were rigorously trained to make reliable
and accurate observations. Encountering many situations of misfortune among

patients, they adapted themselves remarkably well and met the study needs with
objectivity and simultaneous concern for humanistic values. This most capable team

included: nurses Claire B. Balian, Edna W. Duchon, Sylvia M. Bloomenthal, and
Beverly A. Tahari; physicians Reed T. Keller, William B. Newberry, Jr., and
Charles L. Schulman; and sociologists Marjorie W. Jaffe and Mildred G. Sindell.

Dedicated to help the chronically ill people in the study was the staff of the
Visiting Nurse Association of Cleveland. The agency’s full resources were made
available, including its staff education program, trained supervisors, and compre-
hensive services. As the nurse coordinator for the study, Miss Hazel Deuble
conscientiously carried out the project tasks within the agency and served as liaison
between the investigators and the agency personnel. Appreciation is due to Janet R.
Price, Assistant Administrator of the agency, and to all members of the agency’s
staff, because all were directly or indirectly involved at one time or another.
Particular acknowledgment is here expressed to the nurses who served the patients
directly, devoting their skills with compassion to the long-term care of these
chronically ill people.

In a community study as far-reaching as this, many others contributed time and
skills. Though it is not possible to give credit to each and every one by name, sincere
appreciation is nonetheless extended to all. Among these were the aides, nurses,
physicians, social workers, therapists, administrators, and other staff of Abington
House, the hospital where patient selection took place and where professional
referral to the study’s service program was initiated. Trustees of the Benjamin Rose
Institute were responsible for the operation of the hospital, and cooperated fully
during the period of the study. The Cleveland Academy of Medicine endorsed the
study and encouraged cooperation by local physicians who, without exception,

Preface

ix
Effects of Continued Care

responded as requested to the service and investigative needs of the study.

Appropriate acknowledgment to the secretarial, programming, and other technical
staff of the study is difficult to express. Contributing many woman-years of work
with data and manuscripts, theirs was one of the indispensable elements of the
project’s activities, and is gratefully acknowledged here. Lastly, we acknowledge that
this publication is based in part on work performed pursuant to contract HSM
110-69-240 from the National Center for Health Services Research and Develop-
ment, HSMHA, DHEW; and in part on support in the form of research grants
NU-00067, HD-00669, and GM-12302.
 

CHAPTER I.

The contribution which a nation makes to
human history may be judged not only by how
it uses economic and political power, but also by
how it treats its least favored citizens. The social
upheaval following the Second World War has
forced the people and the government of the
United States to recognize important changes in
the social structure of the country. Fortunately,
during the past twenty years, an unprecedented
economic boom has made possible some ad-
vancements for the underprivileged, in spite of
two costly wars.

During the same two decades interest has
been attracted, less dramatically but no less
urgently, to the needs of those who are included
in the widening span of social dependency.
Three out of every ten Americans are now in
school. The problems of youth and education,
though still far from solution, have received
increased attention which has led, among other
things, to our largest and most rapidly growing
program of public assistance (1).

In 1965, public attention was directed to the
increasing numbers of citizens disabled by age
and illness. The passage of the Medicare amend-
ment to the Social Security Act marked a
turning point in the history of social legislation
in this country. It recognizes formally the
obligation of the producing members of society
to meet the great needs for health services
experienced by those who can no longer earn
their own living. The implications of this act for
the future are many. Most obvious is the fact
that it is bringing about a transfer of medical
costs from insurance programs, charities, and
private sources to public taxation. The act also
incorporates standards of service for nursing
homes and hospitals. It opens up new demands

PART |

the plan, elements, and results of the study

INTRODUCTION

by providing funds for groups of people many of
whom had been medically indigent. By linking
payment for medical services to the established
Social Security system, it sets a pattern for the
financing of health services which will almost
certainly expand. The final effect of this legisla-
tion on the amount, character, and distribution
of health services in this country is hard to
predict exactly, but it will surely be great (2, 3).

In commenting on a related development in
health services, namely, the expansion of hos-
pital services, Titmuss points out that ends tend
to be obscured by excessive preoccupation with
means (4). The same danger is inherent in a
welfare program of the scope of Medicare. The
economic purpose of the program is clear. A
long public and legislative debate has culminated
in general agreement that, since private insur-
ance cannot sustain the cost of caring for the
aged and permanently disabled, this economic
burden must be shared by all taxpayers. Execu-
tion of the program, encumbered as it is with
safeguards and compromise provisions, puts new
administrative burdens on every segment of a
loosely organized health care system. The goal
of better health at first seems buried under an
avalanche of paperwork, utilization review com-
mittees, vendor payment provisions, and con-
fused expressions of demands and needs from
the aged and from the professionals who serve
them. In the final analysis, the impact of the
program on the financing of health services will
be difficult to determine, and the effect which it
is intended to have on the health and welfare of
the people will be even more difficult to
measure.

Medicare, thus, raises in a specific area the
general question of how health services can be

1
Effects of Continued Care

evaluated. In broad terms, of course, the answer
is: by scientific measurement and objective
comparisons. The term ‘health services”, how-
ever, opens up an extensive field in which
medicine, public health, ethics, political econ-
omy, and other interests all play a part. Ade-
quate study of even a small element of the
health care system requires much time, many
skills, a variety of points of view, and careful,
critical efforts at quantitation. The investigation
which we present in the following pages illus-
trates the problems encountered in the evalu-
ation of one element of the system, namely, the
use of home nursing services to maintain or
improve the function of chronically ill, elderly
people. In order to represent fairly and objec-
tively as many of the ramifications of such a
program as possible, nurses, doctors, sociolo-
gists, psychologists, and biostatisticians col-
laborated in the study. Details of planning,
executing, and analyzing a project in which data
collection alone took three-and-a-half years were
inevitably complex.

The main outlines and findings of the study
are presented in Part I of this book. In Part II,
we give special consideration to the methods.
Before proceeding to the study itself, however,
we shall place the problem in its proper context
by sketching the main features of chronic illness
in the United States today and by summarizing
published material about home health services
after rehabilitation. We shall complete the chap-
ter by introducing the research questions and
approach.

Chronic Disease in American Society

One in ten Americans is now 65 years old or

older. As in other developed countries, this
proportion is higher than ever before. The

present population structure is the result of high
fertility and immigration before 1900, and
decreasing death rates since. Special prominence
has been given to the current population of aged
people by the succeeding lower fertility of the
depression years. Current higher birth rates, if
continued, will eventually lead to a more evenly
graduated population pyramid, but there is no
reason to expect the numbers of old people to
decrease. For many of the aged, chronic illness

and disability are a way of life, and the
prevalence of most chronic illnesses increases
rapidly with age. Most people over 65 report
that they have one or more ‘chronic condi-
tions”. One-third say that they have limited
mobility, and more than half describe them-
selves as limited in a major activity, such as work
or housekeeping (5). A continuation of the
current population increase will inevitably lead
to large increases in the numbers of people with
chronic illness and disability (6).

Mortality figures are often used by those who
are concerned with a health problem such as
coronary disease to support claims for public
attention and for funds. Less striking, but more
directly expensive, are the residuals left by the
major chronic diseases as, for example, the
progressive restrictions of rheumatoid arthritis
or the incomplete catastrophe of a stroke.
Prevalence of disability is more difficult to
measure than death rates. This fact and the
relatively low social value which our culture
assigns to old people help to explain why their
disabilities have been given less attention than
the more impressive causes of death among
young adults. Among the leading causes of
disability, heart disease ranks first as a cause of
activity limitation in the United States, accord-
ing to people interviewed in their homes (5).
Arthritis and rheumatism rank second, and

‘impairments of the back and spine third. Among

patients in nursing homes, stroke is the most
common condition, followed by heart disease
and arthritis and rheumatism (7). The mental
and behavioral deterioration variously desig-
nated chronic brain syndrome, senility, senile
dementia, or, euphemistically, “mental and nerv-
ous conditions”, also ranks high in any survey of
disability, although measurement is least precise
here.

A particularly knotty semantic problem is
that of distinguishing among the overlapping
terms, disease, illness, and disability. We use the
word “disease” to refer to the presence of any
morbid physical or mental state, latent or
manifest. “Illness refers to disease with mani-

fest clinical symptoms, while “disability” means
demonstrable loss or limitation of normal func-

tion from whatever cause. Thus, for example,

\
while 16 to 20 per cent of the population in one
urban study had some evidence of heart disease
by clinical examination, 7 per cent reported that
they were ill because of heart disease, while only
1.5 per cent were judged to have moderate or
severe disability due to heart disease (8).

The total number of disabled persons in the
United States amounts to at least 3.5 per cent of
the population, or 7,300,000 people. About
600,000 are in mental hospitals (9), and a
similar number are in institutions for the aged
and chronically ill (7). Smaller groups of handi-
capped children are in special schools or institu-
tions. By far the largest group, however, consists
of the 6.1 million persons who live at home even
though they report some degree of limitation in
their ability to move about (5).

The traditional public response to the pres-
ence of age and disability in the community has
been to establish special institutions such as
“homes” for the aged, asylums, nursing homes,
and chronic and mental disease hospitals. Two
linked social motives, one of removing an
inconvenience and the other of providing special
care, have influenced the character of every such
organization. As long as they have existed, these
institutions have had their critics and reformers.
Partly as a reaction to the multiplication of
institutions in the nineteenth century, criticism
of public and private “total” institutions has
become more explicit in recent years, emphasiz-
ing the dangers of bureaucratic impersonality
and the potentially harmful psychological ef-
fects of the institution on its members (10,
11,12). This criticism has contributed to a
resistance to the use of institutions, with an
inevitable increase in the already large numbers
of disabled who live in the community (13).

Few if any communities in the United States
are prepared to meet the simplest needs of the
existing population of chronically ill persons
who live at home. Social organizations which
once bore the weight of responsibility for these
needs are less able to provide care for the old
and disabled. The family unit is pulled apart by
occupational mobility, intergenerational con-
flict, and the costs and pressures of urban living.
Churches are beset by financial problems and
loss of authority. Voluntary charities are limited

Introduction

in the support that they command, and their
charters may not permit them to cope ade-
quately with modern social needs. At the same
time, public expectations of health services are
increasing, and the availability of general practi-
tioners to meet expectations and needs is de-
creasing. Strained by increased numbers of
elderly and chronically ill persons discharged
from mental hospitals and other institutions, the
existing methods of providing care for the
chronically ill and disabled at home show up
repeatedly as some of the weakest links in our
health care system (14).

In contrast with the development and refine-
ment of in-hospital services in the last two
decades, the inventory of services outside the
hospital is meager and out of date. This imbal-
ance has arisen partly from a great increase in
technological skills and methods, with a corre-
sponding proliferation of specialties which can
hardly be practiced outside a modern hospital.
In the United States, the trend toward central-
ized services in the hospital or health center has
been accelerated by the rapid growth of health
insurance which rarely covers out-of-hospital
services. In spite of vigorous efforts at demon-
stration and innovation, the development of
ambulatory clinics, day care centers, home care,
and homemaking services has been slow and
limited in comparison with those in Scandinavia
and Great Britain (15,16). Increasingly, the
patient is expected to come to the medical
center for access to scientific medicine and
scarce professional time, yet the development of
ambulance and transfer services has not kept
pace. For the disabled patient confined to his
home, modern medicine tends to become more
remote and his contact with it more episodic
and haphazard. Clearly, he needs help.

Home Health Services After Rehabilitation

The Commission on Chronic Illness recom-
mended in 1956 that home care be expanded as
an alternative to institutionalization of the
long-term patient (17). The report also suggested
that public health nursing was probably the
most commonly used and perhaps the most
important element in a successful program of
home care. Public health nursing holds a special

3
Effects of Continued Care

place among the health professions in that it
directs its services to patients in the community
and, particularly, to those at home. The public
health nurse is an essential part of home care
programs, whether they are based in hospitals or
in the community. Home nurses employed by
county and state health agencies and by inde-
pendent visiting nurse associations, in fact,
furnish a significant amount of the home health
services now available to the American public.
The Medicare act, further, recognizes her im-
portance by providing special payment for her
services. Despite these facts, inadequate atten-
tion has been paid to the nature and effective-
ness of these services. We must take stock of our
present resources before we can strengthen and

add to them; hence, we return to the funda-
mental question: how effective are the services

we are now using?

At the time our project was undertaken,
review of the published literature on the aged
chronically ill revealed no controlled study of
the impact of health services on patients’ prog-
ress following discharge from a hospital rehabi-
litation program. Some pertinent reports were,
however, available on such related topics as
chronic illness, home care, and medical and
public health nursing supervision of the chroni-
cally ill.

One study, other than ours, was known to be
in progress and was also concerned with the role
of the public health nurse in the follow-up of
chronically ill patients after their discharge from
the hospital. This was the study of continued
care of impaired older persons by the Com-
munity Service Society of New York. It dealt
specifically with disabled, older individuals who
had received rehabilitation services and had been
discharged to the community (18, 19). During
the planning stage of the presently reported
experiment, it seemed that the two studies
would be complementary, since they had related
interests and dealt with patients of contrasting
social and economic backgrounds.

The Associated Hospital Service of New York,
in a study conducted from 1952 to 1957,
concluded that at some point in their hospital-
ization 10 per cent of medical and surgical
patients became candidates for home care in-

stead of continued hospital care (20). This
evidence of reduced number of days of hospital
stay and decreased costs had resulted in the
amendment of the New York State Insurance
Law to provide coverage for care in the patient’s
home. Up to 1960, most studies of home care,
as the one just cited, had been directed to the
economic question of whether home care could
be used to reduce hospital costs. Our study, in
contrast, was designed to assess the value of
home nursing services in maintaining the gains
of a hospital rehabilitation program rather than
as a means for reducing costs.

A final type of relevant study had to do with
continuity of services between general hospitals
and public health nursing agencies (21). As one
important judgment, the author recommended
that a systematic study of continuity of patient
care be made in order to determine ‘““‘the effect
of referral and public health nursing services in
terms of improvement of patients’ conditions
and prevention of progress of disease”.

The Research Approach and Research Questions

In analyzing the methods available for assessing
medical care provided by physicians, Dona-
bedian identifies three possible approaches (22).
The most indirect is the study of the settings
and instrumentalities of care, on the question-
able assumption that good resources necessarily
produce good results. Closer to the point is the
second method, the study of the process of care,
with judgments based on defined values and
standards. Using this method, description may
be accurate, but a large subjective element of
judgment cannot be avoided. Judgments often
vary, and such a study tends to reveal more
about the raters than about the service being
rated. The most direct method of assessing
medical care is to use objective measures of
outcome as the criteria of quality. Preciseness
and objectivity permit more reliable inferences;
therefore, the decision was made to use outcome
criteria in the present study. As Donabedian also
points out, the kind of outcome must be
specified; and he lists three kinds, namely,
recovery, restoration of function, and survival.
Recovery cannot be used in the study of chronic
diseases which are incurable. In the present
study, restoration or maintenance of function
and survival were assessed; and an additional
aspect of outcome was examined, namely, use of
services.

Another problem to consider in the use of
outcome criteria is that of the relevance of the
measures to the services under study. Can
periodic home visits by a nurse, for example,
reasonably be expected to alter survival? Main-
tenance of function appears to be a more
relevant criterion of nursing service. Is increased
use of services to be interpreted as costly excess,
necessary therapy, or preventive medicine? Such
questions can become quite involved and their
interpretation correspondingly uncertain.

Home care of the chronically ill takes place
“in the field”, as the epidemiologist would say,
and it is at least as much a social process as a
medical one. Epidemiological and sociological
methods must, therefore, be combined with
clinical observation; and because of this com-
plexity, most studies of health services tend to
be limited to description. Case studies are
sometimes used to evaluate services and to offer
a framework, although a shaky one, for compari-
sons and decisions about services. Rarely does
the opportunity arise to evaluate a program of
health services by establishing a control group
according to strict methods of random assign-
ment. The present study was designed to take
advantage of one such opportunity.

As described in the Preface, a favorable set of
local circumstances made the study possible. We
were physicians, nurses, and biostatisticians who
had long-standing professional experience with
chronic illness; and we were engaged in teaching,
in research, and in hospital and home care.
Having practical and theoretic interests in com-
mon, we collaborated to plan and to conduct an
experimental, controlled study of the outcome
of a sustained program of home nursing services
for patients discharged from a chronic disease
rehabilitation hospital. The outcome would be
defined by measured changes in the well-being
of experimental and control patients.

We agreed that it would be desirable to
hypothesize detailed relationships between spe-
cific elements of home nursing care and relevant
measures of outcome. A nurse’s supervision of

Introduction

prescribed exercises, for example, might result in
improved physical mobility; but many other
factors would also influence mobility over a
period of time. Some factors would be indirectly
related to the nurse’s activity, such as general
health and morale; and some would be largely
unrelated, such as the natural progression of the
disease process or the presence of physical
obstacles such as stairs. Because of the multi-
plicity and complexity of relationships between
the experimental variable and measures of out-
come, we decided to state a single, general
hypothesis and then to examine it from as many
different specific viewpoints as possible. The

hypothesis was that:

after discharge from a chronic disease rehabilita-
tion hospital, patients whose care is regularly
supervised in the home by a public health nurse
(working with the patient’s physician) will more
often maintain or increase physical, psychological,
or social function than will patients whose care is
not so supervised.

A second hypothesis was implicit in the
planning of the study. This hypothesis was based
on the concept that home care by public helath
nurses is one link in a chain of health and
welfare services (23). Viewed in this way, one of
the important effects of the nurse’s services is to
facilitate the patient’s access to other health
services and to increase their use for appropriate
purposes. This is not a new idea, but it assumes
new importance as the complexity of health
services increases. Public health nurses have, for
decades, been urging patients to see their doc-
tors and to keep clinic appointments. Like other
members of the health service team, the public
health nurse functions as an essential “enabler”
(24). Our concern with this aspect of public
health nursing was evidenced by the fact that
the research protocol included questions about
the use of hospitals and nursing homes, and
included counts of the services of doctors,
nurses, and other health professionals. The
second general hypothesis of the study was,

therefore, that:
after discharge from a chronic disease rehabilita-
tion hospital, patients whose care is regularly
supervised in the home by a public health nurse
(working with the patient’s physician) will make
more use of other health services than will patients
whose care is not so supervised.
Effects of Continued Care

The testing of these two hypotheses was the
immediate objective of the research. In under-
taking this complex and time-consuming study,
llowever, we intended it to have practical as well
as theoretical value. Although we examined only
a part of the health services system, it is a part of
the system about which little is known, in spite
of a pressing need for development. The descrip-
tion of outcome in relation to services given to a
large group of chronically ill patients could, in
itself, provide useful guidance for practitioners
and teachers of public health nursing, as well as
for physicians who are responsible for the care
of the chronically ill at home. The scope and
duration of the study would offer opportunities
to develop and test methods of measurement
and longitudinal data analysis. Above all, spe-
cific knowledge about predictable results of
home care would add to the basis of fact on
which rational planning must rest. Finally, we
hoped that a study focused on the welfare of the
patient would draw attention to a neglected
aspect of health services, with resulting benefit
(0 those who are disabled by age and chronic
illness.

References

I. Clark, D. W., and MacMahon, B.: Preventive Medi-
cine. Brown and Co., Boston, Mass., p. 801, 1967.

2. Somers, H. M. and Somers, A. R.: Medicare and
the Hospitals: Issues and Prospects. Published by
the Brookings Inst., Washington, D.C., pp. 255-291,
1967.

3. Roemer, M. 1.: The Future of Social Medicine in the
United States. Published by Edutext Publications
Ltd., London, England, pp. 1-16, 1967.

4. Titmuss, R. M.: Essays on ‘The Welfare State’.
Published by Yale University Press, New Haven,
Conn., pp. 146, 150, 210, 1959.

5. National Center for Health Statistics: Limitation of
Activity and Mobility Due to Chronic Conditions,
United States, July 1965-June 1966. Series 10, No.
45. Published by the U.S. Department of Health,
Education, and Welfare, Public Health Service,
Washington, D.C., pp. 1-8, May 1968.

6. Lilienfeld, A. M. and Gifford, A. J.: Chronic
Diseases and Public Health. Published by The Johns
Hopkins Press, Baltimore, Md., pp. 5-28, 1966.

7. National Center for Health Statistics: Prevalence of

10.

11.

12.

13.

14.

15.

16.

17.

Chronic Conditions and Impairments Among Resi-
dents of Nursing and Personal Care Homes, United
States, May-June 1964. Series 12, No. 8. Published
by the U.S. Department of Health, Education, and
Welfare, Public Health Service, Washington, D.C.,
pp. 12-22, July 1967.

. Commission on Chronic Illness: Chronic Illness in

the United States, Vol. IV, Chronic Illness in a
Large City. Published for the Commonwealth Fund,
Harvard University Press, Cambridge, Mass., pp.
76-82, 1957.

. National Center for Health Statistics: Characteristics

of Patients in Mental Hospitals, United States,
April-June 1963. Series 12, No. 3. Published by the
U.S. Department of Health, Education, and Welfare,
Public Health Service, Washington, D.C., pp. 14-17,
December 1965.

Commission on Chronic Illness: Chronic Illness in
the United States, Vol. 11, Care of the Long-term
Patient. Published for the Commonwealth Fund,
Harvard University Press, Cambridge, Mass., pp
163-235, 1956.

Townsend, P.: The Last Refuge: A Survey of
Residential Institutions and Homes for the Aged in
England and Wales. Published by Routledge and
Kegan Paul Ltd., London, England, pp. 338-370,
1962.

Goffman, E.: Asylums. Essays on the Social Situa-

tion of Mental Patients and Other Inmates. Pub-
lished by the Aldine Publishing Co., Chicago, Ill,
pp. 1-124, 1962.

Tizard, J.: The Integration of the Handicapped in
Society. Published by Edutext Publications Ltd.,
London, England, pp. 3-8, 1966.

National Commission on Community Health Serv-
ices: Health is a Community Affair. Published by
the Harvard University Press, Cambridge, Mass., pp.
30-31, 1967.

Silver, G. A.: Family Medical Care: a Report on the
Family Health Maintenance Demonstration. Pub-
lished by the Harvard University Press, Cambridge,
Mass., pp. 123-152, 1963.

Ryder, C. F. and Frank, B.: Coordinated Home Care
Programs in Community Health Agencies—A Decade
of Progress. Am. J. Pub. Health, 57:261-265,
February 1967.

Commission on Chronic Illness: Chronic Iliness in
the United States, Vol. 11, Care of the Long-term
Patient. Published for the Commonwealth Fund,
Harvard University Press, Cambridge, Mass., pp
5-81, 1956.
18.

Posman, H., Kogan, L. S., LeMat, A. P. and Dahlin,
B.: Continuity in Care for Impaired Older Persons.
Public Health Nursing in a Geriatric Rehabilitation
Maintenance Program. Report Prepared for the
Committee on Health by Harry Posman, Depart-
ment of Public Affairs, Community Service Society
of New York, 105 East 22 Street, N.Y., N.Y.,
126-148, December 1964.

. Dahlin, B.: Service Following Rehabilitation Train-

ing. A Description of Public Health Nursing Service
Offered a Group of Disabled Older Persons. Pre-
pared for the Committee on Health, a Supple-
mentary Report (see Posman, Continuity, etc.)
Department of Public Affairs, Community Service
Society of New York, 105 East 22 Street, N.Y.,
N.Y. 47-67, July 1965.

20.

21.

22.

23.

Introduction

Associated Hospital Service of New York: Home
Care Following Hospitalization. First Progress Re-
port, published by the Associated Hospital Service
of New York, N. Y., N.Y, 85-103, 1961.

Smith, L. C.: Factors Influencing the Continuity of
Nursing Service, Published by the Columbia Univer-
sity Press, N. Y., N.Y., pp. 1-6 and 85-103, 1962.

Donabedian, A. D.: Evaluating the Quality of
Medical Care. Published by the Milbank Mem. Fund
Quart., 44 (suppl.), pp. 166-206, July 1966.

National Commission on Community Health Serv-
ices: Health is a Community Affair. Published by
the Harvard University Press, Cambridge, Mass., pp.
17-37, 1967.
CHAPTER II.

Careful planning of the research was essential to
avoid predictable sources of bias. An under-
standing of the study and its results requires
thorough description of the plan and process, as
well as detailed definitions of terms. Needed,
also, are descriptions of the two kinds of health
service programs which are identified in our
hypotheses, namely, the programs of the chronic
disease rehabilitation hospital and of the volun-
tary community agency which provided home
nursing services. We describe the research plan
and its implementation in this chapter. In the
next three chapters, we present information
about the patient sample, the service programs,
and the study measures.

Practical problems often set stringent limits
on how one can study a program of health
services by experimental methods. For example,
potential subjects and service personnel may
object to the randomized giving and withholding
of service if it seems to them unnecessary or
unfair. A second type of limitation is the
difficulty of accumulating complete and repro-
ducible information in long-term prospective
studies, a limitation which is particularly evident
in the field of chronic illness where patients may
be lost through death or other causes and where
the research program is difficult to sustain. A
third problem is that of finding measures which
are both reliable and valid and which can serve
as relevant criteria of the outcome of treatment.
On the other hand, nonexperimental studies are
always open to the criticism that the cause and
effect relationships between the health services
and the study results are rarely clear.

The circumstances of the present study per-
mitted us to overcome certain of these limita-
tions and to plan a prospective, controlled,
experimental study based on random assignment
of treatment. Favorable elements were: (1) two
of us, as directors of the hospital and the public

THE RESEARCH PLAN

health nursing service involved, could regulate
important aspects of the service program, and
(2) our research team was experienced in observ-
ing the progress of patients with chronic illness.

We defined the “treatment” to be studied as
the program of care which resulted when a
public health nurse was assigned to assist in the
care of a patient discharged from a hospital for
the chronically ill. This one variable was to be
controlled (i.e., provided for or withheld from
study subjects) as carefully as possible. Other
variables which might affect the outcome, such
as age, diagnosis, and functional state, were

considered to be too numerous to serve as a
basis for stratified sampling. Certain variables, in

fact, which might affect the outcome could not
be identified. We, thus, chose a system of
random assignment as the method of allocating
patients to the experimental groups, with a
sample size large enough (300 patients) to allow
comparable distribution of extraneous variables
in each experimental group.

Research Design

The basic plan of the study was that 300
patients were randomly assigned to treatment or
control groups at the time of discharge from the
hospital, the treatment consisting of regular
visits by a public health nurse to the patient at
home. Each patient was observed during a
period of two years after discharge by a separate
team of observers trained to apply specific
measures of physical, psychological and social
function. We separated observers from care
personnel in order to avoid the bias of those
who gave service. In addition, we regulated the
frequency of research observation. Patients were
randomly assigned either to a group observed at
quarterly intervals or to a group observed only
at the beginning and end of the study, in order
to permit detection of effects related to the
process of observation.
We designed the sample size as large as
available resources and time would permit, so
that subgroups within the somewhat heterogene-
ous total groups could be examined. We chose
the observation period of two years because the
illnesses which originally led to admission to this
particular hospital were largely chronic and
usually necessitated a long-term nursing plan.
Previous studies had shown that, although meas-
urable improvement in function occurred most
frequently during the first year after discharge,
some patients improved between the first and
second years (1, 2). Observations were not ex-
tended beyond two years because two-year
mortalities of the order of 40 per cent had been
observed in patients with stroke and hip fracture
discharged from this hospital. This experience
suggested that the sample could become too
small to be useful after two years. Three months
was selected as the interval between observa-
tions. We estimated this to be a period long
enough to permit the available staff to visit the
assigned patients in their homes and yet short
enough to observe and recall changes in func-
tional status.

A well-recognized principle of behavioral re-
search is that the process of observation influ-
ences the behavior itself (3). Since we planned
to observe the subjects at intervals for two years,
we could predict that a relationship would
develop between the observer and the patient.
Such contacts, which would in certain ways
resemble the interaction between the public
health nurse and her patient, might affect the
patient’s behavior and obscure differences be-
tween treated and control groups. For example,
a patient repeatedly questioned about when he
last saw the doctor and about his functional
limitations might be prompted to make more
frequent appointments with his doctor, and
might change his approach to his physical
limitations, whether the questions were asked by
a neutral observer for research purposes or by a
public health nurse undertaking a plan of care.

We adopted, therefore, the sample design
presented in Table 2.1 in order to permit identi-
fication of changes related to the nursing pro-
gram, changes related to regularly scheduled
observation, and changes related to interactions

The Research Plan

Table 2.1—-Basic Design of Sample*

 

 

 

Referred Not
referred
to Visiting LL Totals
Nurse (N+) to Visiting
Nurse (N-)
Observation
classes
Observed
every three
months (0+) 75 (N+O+) | 75 (N-O+)| 150
Observed
only at
intake and
termination (O-) | 75 (N+O-)| 75(N-0-)| 150
Totals 150 150 300

 

 

 

 

 

 

*Figures are numbers of subjects.

between the nursing program and observation
process. The design is called a factorial design,
and methods for analyzing and interpreting
results based on such designs are well known
4,5).

We established four equal-sized groups which
differed systematically both in the treatment
assigned and in the frequency of observation. By
this means, we distinguished several types of
effects. Effects on patients who were assigned to
the nursing program were identified by compar-
ing the 150 patients who were “Referred to
Visiting Nurse” (Table 2.1) with the 150 who
were “Not Referred to Visiting Nurse”. In this
comparison, effects of the observation process
alone were eliminated since the presence or
absence of systematically repeated observation
was balanced equally between the groups as-
signed and not assigned to the nursing program.

Effects related to systematically repeated
contacts with an observer were recognized by
comparing the 150 patients who were “Observed
every three months” (Table 2.1) with the 150
who were “Observed only at intake and termina-
tion”. In this comparison, effects of the nursing
program alone were eliminated since the pres-
ence or absence of the nursing program was
Effects of Continued Care

balanced equally between the groups assigned to
systematically repeated observation.

Effects on patients who were assigned to the
nursing program (comparison of the 150 who
were “Referred” with the 150 who were “Not
Referred”) were not, in all instances, at-
tributable solely to the nursing programs, since
interaction between the nursing program and the
observation process was a potential source of
bias in one-fourth of the sample and was not
balanced in the study design. We could, how-
ever, measure and sort out the interaction bias.
Differences related to interactions between the
nursing program and observation process were
identified by comparing combined groups in
which the nursing program and systematically
repeated observation were present separately
(N+O- and N-Ot+ groups in Table 2.1) with
combined groups in which the nursing program
and systematically repeated observations were
both present and both absent (N+O+ and N-O-
groups in Table 2.1).

In some instances, e.g., analyses of quarterly
interval findings, it was necessary to examine
nursing effects by comparisons between the 75
systematically observed patients who were “Re-
ferred to Visiting Nurse” and the 75 systemati-
cally observed patients who were “Not Referred
to Visiting Nurse”. In such instances, the obser-
vation process was a potential source of bias for
interpretations about nursing effects. Unless
otherwise stated, nursing program effects re-
ported in the book are effects that were neither
observer-biased nor interaction-biased.

Study Intake

According to the research plan, we admitted 308
patients into the study at the time of discharge
from Abington House (formerly the Benjamin
Rose Hospital). The 308 were those consecutive
patients who met the following criteria: (1)
discharge to a home (noninstitutional residence),
(2) residence within the area served by the
Visiting Nurse Association, (3) age 50 years old
or older, (4) hospital stay of at least one week,
and (5) did not leave the hospital against advice.
The first eight patients were designated as a pilot
group. The purpose of the pilot group was to
provide an opportunity at each stage of the

10

study for practice and adjustment of procedures.
Data related to these patients were not to be
included in the analyses. Like the other patients,
they passed through the complete 2-year study
process. None of the observers or nurses was
informed of the special status of this group.
Random assignment of patients to observer
and treatment groups took place in two stages.
The biostatistician prepared a set of 308 scaled
envelopes, each envelope containing an observer
assignment and, in an inner sealed envelope, an
assignment to one of the four experimental
groups (Table 2.1). A member of our research
staff kept a log of all patients in the hospital and
recorded the estimated dates of discharge at
regularly scheduled intervals. When a patient
met the intake criteria and when his discharge
was estimated to be seven days away, we
assigned him a study number and opened the

first envelope to determine to which observer he
was assigned. In order to ensure that knowledge

of a patient’s experimental group did not influ-
ence the observer’s initial evaluation or the
hospital discharge plan, we withheld opening the
second envelope which contained the experi-
mental group assignment until an estimated
three days before discharge. We considered three
days to be the minimum required to make an
orderly referral of the patient to the Visiting
Nurse Association. The date on which the
patient was assigned to an experimental group
was designated the intake date. Time intervals
reported later are measured from this date,
unless otherwise specified. Terminal evaluations
of all surviving patients were scheduled on the
second anniversary of the intake date.

The randomization procedure was designed to
ensure that, at the time of intake, each patient
had an equal chance of being assigned to any
one of the four experimental groups. Since we
expected the intake process to last more than a
year (actually lasted 22 months), the scheme of
randomization was divided into blocks of twenty
successive patients, starting with case number 9,
after the pilot group. Within each block, each
experimental group included five patients. By
this means, variations related to time, such as
seasonal effects or changes in service staff
and program, were evenly distributed; and
»

convenient termination points were provided.
We did not inform observers about the details of
the random process.

As soon as a patient was assigned to an
observer (at the time of opening of the first
envelope), the observer introduced herself to the
patient and conducted the intake interview. One
of the physician investigators supplemented this
interview with additional interview, physical
examination, and laboratory data. We explained
the study as an effort to learn more about
patients and their progress in order to improve
patient care.

When the patient was assigned to an experi-
mental group (at the time of opening of the
second envelope), nothing further was done in
relation to those who were to be control
patients (not referred to visiting nurse) except
that the observers made arrangements to see
those who were to be observed regularly. When a
patient was assigned to a treatment group
(referred to visiting nurse), one of the physician
investigators immediately approached the physi-
cian who would be caring for the patient after
discharge and asked his permission to refer ‘the
patient to the Visiting Nurse Association. No
physicians refused this permission. The investi-
gator then initiated the referral, explaining to
the patient and his family that this was recom-
mended by the hospital staff as a means of
maintaining progress achieved in the hospital.

In regard to fees, the usual agency policy was
followed. Each nurse thus evaluated the needs
and resources of each patient during the first
few visits and set a fee, ranging from full cost
($6.25 a visit) to no charge. At the time of
referral to the Visiting Nurse Association, the
members of the hospital staff who had been
caring for the patient completed a standard
form, the Greater Cleveland Confidential Refer-
ral Form, which presented details of diagnosis,
medications, orders for treatment, and sum-
maries of nursing, social work, physical and
occupational therapy experience. This form was
sent to the Visiting Nurse Association.

The care offered to the patients in the
treatment group was basically that regularly
provided by the Cleveland Visiting Nurse Asso-
ciation. We describe nurse selection and neces-

The Research Plan

sary modifications in the patterns of service in
Chapter 4. Although it would have been d sir
able to randomize the assignment of nurses, this
was not practicable.

We carried out the intake process as planned.
In the final analysis, there were 75 patients in
each of the four experimental groups. Of the
300 subjects, all but six had their initial evalu-
ations by the assigned observer. These six were
seen by substitute observers.

Interval Observations

Previous studies of patients discharged from
Abington House had shown that important
changes in such functions as walking and aci'v-
ities of daily living frequently occurred immedi-
ately after discharge and less frequently later.
For this reason it seemed advisable to plan to see
the patient soon after his return home. A second
reason for planning an early visit had to do with
methods. Previous experience with the Index of
Independence in Activities of Daily Living had
brought out the fact that, in addition to valid
changes in function associated with the move
from hospital to home, artifacts might be
present in either setting. The design and location
of closets and bathrooms, for example, might
influence the degree of independence in dressing
and using the toilet. Early observations in the
home were, thus, desirable for us to identify the
number of such artifacts and to establish ©
baseline in the home as well as hospital.

The first set of observations for the 150
patients who were to be visited every three
months was, therefore, scheduled for one weck
after discharge. The date of this post-discharg:
visit is the only one which we did not base
directly on the study intake date. Instead.
because of the logical relationship of this visit to
the actual discharge, we based it on the dis-
charge date. When all the subjects had bee:
taken into the study, we found that the actua
mean interval from intake date to discharge was
5.4 + 7.3 days! where three days had been

 

IThe notation 5.4 + 7.3 days indicates that the average time
between intake date and hospital discharge was 5.4 days, while
the standard deviation was 7.3 days. The distribution of times
here is highly skewed, as evidenced by the fact that the stand-
ard deviation is greater than the mean.
Effects of Continued Care

planned. The post-discharge visit, on the average,
took place 7.2 + 7.1 days after discharge from
the hospital.

For the half of the subjects who were to be
observed quarterly, we scheduled visits (after the
one week post-discharge visit) every three calen-
dar months from the intake date. Visits were
continued by the same interviewer who had
conducted the intake interview until the end of
the planned two-year period of observation or as
long as the patient lived. Observers travelled to
the patient’s home or other residence. The
subjects, as we anticipated, proved to be a
geographically stable group. Only nine moved
out of the community. Of these, seven were
close enough to be visited in person. One moved
to Texas and one to Florida. Information about
these two was obtained by telephone from the
subjects and their personal physicians.

One-half of the subjects, by design, were not
to be interviewed by observers between intake
and the terminal date two years later. One-half
of these were to receive nursing visits, leaving
one-quarter of all the subjects with whom no
contact would take place during the two years.
In order to keep track of this one-quarter, one
of our research staff made a telephone call every
six months to the last known address in order to
determine if the subject was still at the same
address. The regular observers were not in-
formed of this contact.

Deaths

Information that a regularly observed patient
had died was obtained by the observer when she
attempted to arrange the next quarterly inter-
view, and sometimes sooner through obituary
notices or other channels. In the group observed
only at intake and termination, death was
ascertained from the nurses or by routine
semiannual telephone checks and, for all pa-
tients, at the terminal date.

Terminal Observations

The terminal observations were among the most
critical of the study, since it was at this point
that we were to make comparisons among the
four original experimental groups. A primary
consideration was to approach both the ob-

12

served and so-called nonobserved groups in the
same way. For this purpose, the original ob-
server was not appropriate, since she had come
to know half of her subjects quite well over a
two-year span, while she had not seen the other
half since intake. We decided, therefore, to
assign an unfamiliar observer to every subject for
the terminal interview by a random method and
to have the new observer approach every patient
without consulting the regular observer. The
information collected by the new observer is
that which has been used for all group compari-
sons unless stated otherwise. Two physicians
repeated certain evaluations as well. Because of
staff changes, we could not reassign physicians
systematically, but this was not considered to be
essential since they had not seen any of the
patients for the intervening two years.

A second set of terminal interviews was
conducted following the basic ones defined
above. These involved only the quarterly ob-
served subjects and were conducted by the
originally and regularly scheduled observers.
Duplicate terminal data were thus obtained
which permitted us to compare information
obtained by an observer who knew the subject
well with that obtained by one who had not
known the subject previously.

Organization of the Staff

A study which lasts four years and deals with
300 subjects distributed throughout a major
metropolitan area requires a large and well
organized staff. In addition to the nurses who
gave the care, sixteen professional persons, three
secretaries, two programmers, and a keypunch
operator spent major amounts of time on the
study. We established a steering committee
consisting of the principal investigator, the
project director, the biometrician, the directors
of observer and care teams, and a public health
nursing consultant. This group remained the
directing committee from the initial planning
phase through the completion of this book and
was assisted by a psychology consultant.

In an experimental study of health services,
the strict “double blind” technique which is
used in studies of drug effects is frequently not
possible. The purpose of this technique is to
prevent the observers who measure the criteria
from being influenced in their judgments by
knowing which subjects are receiving the experi-
mental treatment and which are the control
subjects. This method is “double blind” since
both subject and observer are prevented from
knowing who belongs in which group. It is
obviously impossible to supply home nursing
care to one group while withholding it from
another and to conceal this difference com-
pletely from subjects and observers. Neverthe-
less, steps can be taken to minimize the bias
from this source.

In specific ways, we designed the program to
prevent observers from being affected by knowl-
edge of who was receiving what service. Prior to
starting intake, the observers were instructed in
the methods of the study. Their headquarters
were remote from the nursing service, and their
only contact with the care team was through
their director (Dr. Mary Adams) who sat on the
steering committee. We did not tell the observers
which patients were receiving nursing service,
and we gave them only general information
about the goals and methods of the nursing
program. We instructed them not to offer any
kind of service and to refer questions about
service to the team director. She, in turn,
communicated such questions to the nurses only
in the rare instances when there seemed to be no
other way to meet a critical need.

The care team consisted of the Visiting Nurse
Association staff, augmented by a nurse coordi-
nator employed for the study. The director of
the agency served as director of the care team.
The nurses were asked to keep special records
and to keep treatment cases open for two years.
We did not give information to agency staff
about which patients were under regular obser-
vation, nor about the kinds of observations
which were being made. Criteria of observer
judgment were not explained to the nurses. We
referred all questions which arose about the
overall study or the observers’ visits to the
director of the care team and, thus, indirectly
back to the observer team. In chapter 4, we
present information more fully about the nurs-
ing program.

In the lives of many of the people who were

The Research Plan

being studied, the visit of the nurse or the
observer was an important event. Subjects occa-
sionally told members of one team about the
activities of the other. Observers also discovered
that some discharged patients maintained regular
contact with friends they had made in the
hospital, even though they lived in different
parts of the city. There was, therefore, a small
amount of inevitable communication between
experimental groups. Communications between
those providing the care and those evaluating
the results were, by design, made sufficiently
difficult that no major communication of this
latter type was possible.

Data Collection

“Physical, psychological, and social function”
are the criteria by which we evaluated experi-
mentally the outcome of the home nursing
program. Many of these functions involve ordi-
nary activities or behavioral responses to a test
stimulus. The standardization and measurement
of such responses is beset by many limitations.
The conditions of observation are variable, and
certain data consist of partially judgmental
reports of human behavior. In view of this, we
used precise and objective measurements when
they were available, as in the definition of
disease states in terms of laboratory findings.

Some of the measures which are most relevant
to our hypotheses, however, depend for their
validity upon skillful interviewing or trained
personal observation. The preparation and train-
ing of the observers and the methods they used
to elicit information was, therefore, an impor-
tant determinant of the quality of the data. The
following general principles are developed in
greater detail in Chapter 5S.

The individuals who collected the data had
training in a variety of professional fields. Three
were physicians. One of the two full-time
observers was a registered nurse, and one had a
master’s degree in sociology. Part-time observers
and substitutes were also nurses and sociologists.
We assembled the research staff and instructed
them in the measures to be used and in
interviewing methods before intake began. Three
observers had had previous experience with
similar studies in the same setting. Throughout

13
Effects of Continued Care

the study, the observers met daily with the
director of the observer team and at intervals
with the principal investigator and project direc-
tor to report their experiences, to clarify defini-
tions, and to obtain guidance in difficult cases.

Concerted efforts were made to obtain as
complete and accurate information as possible.
We considered the primary source of informa-
tion to be the patient. When aphasia, language
barrier, or confusion made this information
questionable, or in the rare instance when the
patient could not be visited personally, we
turned to secondary sources, such as the pa-
tient’s family or his personal physician, and
occasionally to friends, practical nurses, land-
ladies, nursing home staff, and others. When
information seemed doubtful to the observer
because of inconsistency or confusion on the
part of the informant, checks were made with
other informants. For certain data, such as dates
of birth and hospitalization, the observers rou-
tinely checked information obtained from the
subject against public and institutional records.

We wanted the approach to the patients to be
reassuring and to gain their cooperation over a
two-year period. Interviewers arranged their
interviews according to the patient’s tolerance
and wishes. In order to avoid the use of
clipboards and forms, observers memorized the
data schedules and completed them promptly
after leaving the patient. Observers explained the
purpose and nature of the study in general terms
as an effort to “learn more about you and how
you progress so that we can do a better job of
caring for people with similar difficulties in the
future.” The experimental character of the
study and its official relationship to the hospital
were explained appropriately.

We set limits to the intimacy and professional
content of the interviews. Observers were in-
structed to present themselves only as friendly
inquirers and not to give advice about profes-
sional or personal matters. With few exceptions,
the effort to preserve an interested but impartial
approach was successful.

Records were kept on the forms which are
presented in Chapter 13. Once an interview was
recorded, changes could be made only through
the director of the observer team, who was also

14

responsible for reviewing all the data and assur-
ing its completeness.

From the original data sheets, two coders
coded information independently. Each set was
keypunched separately on IBM cards. The cards
were then compared electronically; and the
director of the observer team reconciled all
discrepancies, working with other members of
the staff on definitions and interpretations. The
reconciled cards were repunched and compared
again, and the process repeated until there was
complete agreement between the two sets of
cards.

The nurses who were assigned to patients in
the study kept separate records. From these
records, data summarizing the dates and content
of each visit were abstracted on mark-sense
sheets, punched on cards, and verified.

We collected some kinds of data in duplicate
in order to permit a formal check of reproduci-
bility. For example, two different observers
interviewed the survivors of the quarterly ob-
served groups after two years. There are limita-
tions, however, to the meaning of reproducibil-
ity measured in this way. Many of the data are
time-dependent, and it is impossible to conduct
simultaneous interviews. The mental and physi-
cal state of the patient during the interview, its
location and timing, and many other variables
may influence a measurement or evaluation
result and cannot be standardized in a study
such as this one.

References

1. Katz, S., Ford, A. B., Heiple, K. G. and Newill, V.
A.: Studies of illness in the aged. Recovery after
fracture of the hip. J. Gerontology 19: 258-293,
July 1964.

2. Katz, S., Ford, A. B., Chinn, A. B. and Newill, V.
A.: Prognosis after strokes. Part IV, Long-term
course of 159 patients. Medicine 45: 236-246,
1966.

3. Roethlisberger, F. J. and Dickson, W. J.: Manage-
ment and the Worker, “Hawthorne Study”. Pub-
lished by the Harvard University Press, Cambridge,
Mass., pp. 179-229 and 399-408, 1939.

4. Cochran, W. G. and Cox, G. M.: Experimental
Designs, Second Edition. Published by John Wiley &
Sons, Inc., N. Y., N. Y., pp. 1-44 and 148-181,
1957.
The Research Plan

5. Fisher, R. A.: The Design of Experiments, Seventh Oliver and Boyd, London, England, pp. 1-106 and
Edition. Published by Hafner, N. Y., N. Y., and 163-180, 1960.

15
CHAPTER III.

During the present century, radical changes have
been taking place in the population age struc-
tures of the developed countries and in the
major health problems with which the health
service systems of these countries are con-
fronted. In the United States, persons aged 65
and over are now approaching an unprecedented
ten per cent of the population. The leading
causes of death and disability have shifted from
the acute infectious diseases, with their impact
on the first segment of the life-span, to the
chronic degenerative diseases, which increase
steadily after early adulthood.

In our daily work as physicians, nurses, and
directors of services, we have had thrust upon
our attention the problems generated by the
chronic illnesses and disabilities of older people.
It is clear to the most casual observer that
existing health services in this country cope with
the health problems of such persons in a
discontinuous and fragmented manner. We
designed the present study, therefore, to find
out whether the planned use of one available
method of coordinating and continuing health
services — namely, a program of continuing visits
by a public health nurse, supervised by a
physician — could be shown to improve or
maintain the function of a group of such
persons.

Men and women 50 years old and older, who
were about to be discharged home after being
hospitalized for a chronic illness or disability,
were chosen as the subjects of this study. In
general terms, these patients and their problems
are typical of those being seen in steadily
increasing numbers in hospitals, rehabilitation
services, public health nursing agencies, and at
other points of contact with the health care
systems in the United States today.

We chose this particular sample because of a
common opinion that such patients need more

16

THE PATIENT SAMPLE

and better coordinated care than they usually
receive (1,2). In addition, these individuals
happened to be passing through one channel of
the health care system over which we could
exert enough control to introduce or withhold a
particular element of health service, namely the
home visits of a public health nurse.

There were additional reasons leading to the
choice of this sample. For one thing, on the
basis of previous studies, we already knew a
good deal about the patients in this particular
hospital and could make some predictions about
them, such as that they would sustain a 20 to 40
per cent mortality over a two year period after
leaving the hospital. We had also developed
measures in the earlier studies which would be
useful for describing these patients. Motivation
for the study came from the need to know more
about alternative ways of providing care for such
chronically ill patients, since information about
methods of caring for this growing population
could be used to improve the service programs
of the hospital and the public health nursing
agency and could increase the fund of knowl-
edge for medical and nursing students.

The Hospital

The hospital, Abington House, from which the
sample was obtained is a unique institution. In
many ways it resembles what Phillips has called
the “intermediate” hospital—intermediate, that
is, between short-term general hospital care and
home or long-term institutional care (3). Abing-
ton House (formerly the Benjamin Rose Hos-
pital) was built by the Benjamin Rose Institute
in 1954. The Institute had been endowed in
1908 by Benjamin Rose, a Cleveland philanthro-
pist, “to provide help and sustenance to needy
older persons.” The hospital is a unit of Univer-
sity Hospitals of Cleveland and is staffed by
faculty members of Case Western Reserve Uni-
versity School of Medicine.

The hospital, located on the border between
the inner city and its suburbs, has provided
hospital and rehabilitation services for patients
referred by physicians about equally from Uni-
versity Hospitals and from elsewhere in the
Cleveland area. There are 66 beds. The patients
admitted to the service are similar in many
respects to those admitted to four chronic
diseases hospitals or units described by Gold-
man (4). Typically, patients at Abington House
need restoration of physical function after sus-
taining, for example, a hip fracture or a stroke.
Many patients, also, need comprehensive evalu-
ation and planning as, for example, an elderly
person who has encountered difficulty in manag-
ing on his own. In addition to medical and
nursing services and the consultation and labora-
tory facilities of a large university hospital,
special services are emphasized, including medi-
cal social work and physical, occupational, and
speech therapy. Planning for care after discharge
is stressed, with most patients returning to the
care of private physicians. The patients are
mainly, though not exclusively, elderly. The
average stay is a little over a month, and
two-thirds of the patients are discharged to their
homes in the community.

Sample Selection

Three hundred and eight subjects were selected
from 831 consecutive patients discharged be-
tween July 15, 1963 and May 5, 1965. The
criteria we used for selection were: (1) discharge
to a home (noninstitutional residence), (2)
residence within the area served by the Visiting
Nurse Association, (3) age 50 years old or older,
(4) hospital stay of at least a week, and (5) did
not leave Abington House against medical ad-
vice. We applied these criteria to all patients in
sequence as they approached discharge, and
every patient who met the criteria was taken
into the study.

The largest number of patients excluded from
the study were excluded because they were not
eligible to receive the services of the Visiting
Nurse Association. This agency, by agreement
with local governmental agencies, provided

The Patient Sample

home nursing services in the central metropoli-
tan area and much of the suburbs, while the
remaining areas were served by local public
health authorities. This agency provided nursing
services only in homes, apartments or boarding
houses, and not in hospitals, nursing homes, or
homes for the aged. Elimination of 227 patients
who did not return home and 143 who lived
outside the Visiting Nurse Association area
accounted for almost three quarters of the
exclusions.

The study design called for the elimination of
certain small groups of patients who were
atypical of the population under study and not
numerous enough to comprise a sub-sample.
These were: those under 50 years of age (38
patients), nonwhites (10 patients), and indi-
viduals who left the hospital against medical
advice (4 patients). Finally, 101 patients who
did not meet other criteria for eligibility were
excluded prior to randomization. Included
among the 101 were 38 patients who died in the
hospital, 23 who were already in the study
because of an earlier admission, 24 who were
hospitalized less than 6 days, and 16 who were
discharged earlier than the patient with the final
study number (whose discharge was delayed).

The first eight patients, as noted in Chapter 2,
were the pilot group. One substitution was
drawn from this group and included in the final
analysis. This occurred when patient number 16
left the hospital against medical advice, and
patient number 5 (who had the same random
treatment and observer assignments) was substi-
tuted. We made no other alteration in the
composition of the sample after study numbers
had been assigned. The sample which was finally
analyzed, thus, consisted of the planned 300
patients, or 36 per cent of the patients dis-

charged in 22 months.

The remainder of this section will be devoted
to a description of the physical, psychological,
and social characteristics of the 300 subjects at
the time of their entry into the study. We shall,
also, make comparisons with data from other
studies to give an idea of the representativeness
of the present sample.

17
Effects of Continued Care

Age

The mean age of the patients in the present
study at the time of intake was 72 years, with a
range of 50 to 94 years. The age distribution,
except for the purposeful exclusion of those
under the age of 50, is similar to that reported
for the patients in four other chronic disease
hospitals (4). Three-fourths of the study patients
were 65 years old or older, and persons in this
age range generally constitute about a third of
the case load of home nursing services. A
national survey in 1963, for example, found that
36 per cent of the patients receiving home
nursing from 219 public health agencies across
the country were 65 and older (1).

The Visiting Nurse Association of Cleveland
reported 32 per cent of their patients to be older
than 65 in 1966, while the Associated Hospital
Service of New York gave home care to 5000
individuals, of whom 42 per cent were 65 or
older (2,5). These figures may underestimate
the problem, since a national sample of house-
hold interviews in 1958-9 showed 6.6 persons
per 1000 in the U. S. population to be receiving
personal help or nursing care at home, of whom
58 per cent were 65 or older (6). Since the
Medicare law went into effect in July, 1966, the
proportion of older people referred for home
care has been increasing.

Sex

There were 99 men and 201 women in the
sample. This a ratio of 49.3 men to 100 women,
which is a much greater preponderance of
women than is found among people in the
United States who are 65 and older. In 1965,
the ratio of men to women for this part of the
population was estimated to be 77.6 to 100 (7).
Among patients over 65 who were receiving
home care, a household survey also revealed a
higher ratio of men to women than did the
present study, namely 64.5 to 100 (6).

Disease

The classification of persons on the basis of
disease offers formidable problems and is dis-
cussed further in Chapter 6. For the purposes of
the present study, we used two bases for disease
classification. First, we assigned to each patient

18

a single principal clinical diagnosis or “diagnosis
leading to hospital admission”. This is a classifi-
cation based on illness, or manifest disease
requiring hospital services. In order to arrive at
this diagnosis, two of the physician investigators
read every patient’s entire chart and resolved
any questionable decisions jointly. This process
resulted in the list of diagnoses presented in
Table 3.1. When these diagnoses are compared
with the distribution of diagnoses in studies of
similar populations, it appears that our sample
includes more patients with stroke, fracture of
the femur, and arthritis, and fewer with heart
disease and cancer than are generally encoun-
tered by public health nurses giving home care to
elderly patients (1,2). This distribution clearly
results from the fact that the patients selected
were those discharged from a hospital which
offers physical rehabilitation as one of its
primary services. The kinds of disease seen in
this study resemble more closely those reported
in a related study which also drew its sample
from a rehabilitation-oriented hospital (8).

The leading causes of death in the United
States among persons older than 50 years of age
are, in order of magnitude, diseases of the
circulatory system, neoplasms, diseases of the
nervous system, and accidental injuries (9).
These conditions are numerically well repre-
sented in this study. The two conditions re-
ported by the National Health Survey to ac-
count for approximately 40 per cent of the
chronic limitations found in persons 45 and
older are heart conditions and arthritis and
rheumatism (10). These are also represented
among the larger categories in our present study.

We established a second classification, based
on systematic screening of all patients at intake
for the presence or absence of twelve commonly
occurring chronic disease abnormalities (Table
3.2). Additional information about the defini-
tions for the study of these particular conditions
is presented in Chapters 5, 6, and 13. The list is
not exhaustive, but it does document the great
prevalence of chronic disease in this patient
population. Only one in seven patients was free
of all the twelve conditions which were consid-
ered. More than 60 per cent had two or more
conditions, and four patients had six. The results
Table 3.1. Diagnoses Leading to Hospital Admission for Study Sample

The Patient Sample

 

 

 

 

 

 

ICD Code! Disease classification No. of Per cent
persons of total
330-398 Nervous System and Sense Organs
Stroke 77
Paralysis agitans 6 ; 104 34.7
Other 21
800-999 Injuries—External Causes
Fracture of femur 59
Other fracture of lower limb 12 | 71 23.7
720-749 Bones and Organs of Movement
Rheumatoid arthritis 16
Osteoarthritis 10 ; 43 14.3
Other 17
400-468 Circulatory System
Arteriosclerotic heart disease 15
Vascular disease with amputation 6 27 9.0
Other 6
300-329 Mental, Psychoneurotic and Personality
Chronic brain syndrome 20 77 73
Depression 2
140-239 Neoplasms 13 4.3
530-587 Digestive System 13 4.3
240-289 Allergic, Endocrine, Metabolic and Nutritional 3 1.0
470-527 Respiratory System 1 0.3
290-299 Blood and Blood-forming Organs 2 0.7
590-637 Genito-urinary System 1 0.3
TOTALS 300 99.9

 

 

LCD Code is taken from the International Classification of Diseases, Adapted, published by the U.S. Government Printing Office,

 

Washington, D.C., 1962.

of this screening, compared to the diagnoses for
which patients were hospitalized, reveal more
clearly the sources of disabilities which may be
expected to influence the outcome of any
treatment program. Cardiovascular-renal disease,
for example, was definitely or probably present
in about three-fourths of the patients, although

less than 10 per cent were admitted for this
reason. Although no patient was admitted to the
hospital because of blindness and only one
because of obesity, each of these conditions
complicated efforts to restore independence to
approximately one out of every ten pa-
tients. Conditions which usually imply a poor

19
Effects of Continued Care

Table 3.2—Chronic Disease Abnormalities

 

 

 

 

 

 

Abnormalities No. of
persons
Cardiovascular-renal System
Arteriosclerotic heart disease 125
Vascular heart disease 65
Hypertension 63
Other cardiac abnormalities
(digitalis or EKG abnormalities
of fibrillation, T-wave changes,
or block) 111
Proteinuria 37
Azotemia 20
Metabolic and Nutritional Conditions
Diabetes 44
Obesity 25
Sense Organs
Blindness 36
Locomotor System
Noninflammatory polyarthritis,
past month 33
Inflammatory polyarthritis,
past month 18
Neoplasms
Malignancy present in past
five years 26
Combinations of Above Abnormalities
None of the abnormalities 42 (14.0%)
One abnormality 75 (25.0%)
Two abnormalities 87 (29.0%)
Three abnormalities 56 (18.7%)
Four or more abnormalities 40 (13.3%)

 

prognosis, such as arteriosclerotic heart disease,
malignancy, and chronic renal failure were
shown to afflict many patients, even though
these were often not the admitting diagnoses.
Such conditions limit rehabilitation goals and
must be considered in interpreting the results of
a program of care.

Physical Disability

Disability is a term which refers to loss of
capacity to perform a function. This definition

20

of disability is similar to that used in the
National Health Survey: “Any temporary or
long-term reduction of a person’s activity as a
result of an acute or chronic condition” (6).
Objectively, disability can only be identified
when an individual is required, by request or by
circumstances, to perform a function and fails to
do so. An investigator in a study such as ours
cannot determine whether there is latent capac-
ity which would result in performance if moti-
vation or other circumstances were different.
The measurements of disability presented in
Table 3.3, therefore, are reports of actual per-
formance.

The subjects had many and severe disabilities
at the time of intake into the study. Only 17
persons needed no assistance with their activities
of daily living. A little more than half had full
use of all four limbs, and only 40 were able to
walk without personal assistance or mechanical
support such as a cane. One-third had been
unable to go outside their places of residence
during the two weeks before being admitted to
the hospital. On the other hand, few were
totally disabled. Only four were completely
bedridden, and all but 50 were walking in some
fashion.

Patients with strokes composed the largest
diagnostic group, but other conditions such as
arthritis and multiple sclerosis also produced
limitation of movement in the limbs. As noted
in Table 3.3, we found that 31 per cent of the
300 patients were severely restricted in strength
and range of movement, meaning major limita-
tion in the function of at least one limb (rating
of U3 for an arm or L3 or L4 for a leg as defined
in Chapter 13). The most common pattern was
moderate or severe limitation in one limb (53
patients). Hemiplegia or hemiparesis was identi-
fied in 49 patients, weakness of both legs in 5,
of both arms in 10, of three limbs in 7, and of
four limbs in 11. One patient had limitation of
an arm and a leg on opposite sides.

In summary, most of the subjects exhibited
moderate to severe physical disability in addi-
tion to disease conditions with serious prog-
noses. These disabilities were chronic at least in
the sense that they had not been eliminated in
the course of a hospitalization lasting more than
The Patient Sample

Table 3.3—Physical Disability

 

 

 

 

 

 

 

 

 

 

No. of Per cent
persons of total
Activities of Daily Living!
(Bathing, dressing, toileting, transfer, continence, feeding)
Independent in all 6 activities 17 5.7
Independent in all but 1 or 2 activities 39 13.0
Dependent in 3, 4, or 5 activities 220 73.3
Dependent in all 6 activities 24 8.0
TOTALS 300 100.0
Walking
Walking without personal assistance 83 27.7
Walking with personal assistance 167 55.7
Not walking 50 16.7
TOTALS 300 100.1
Strength and Range of Movement!
Minimal or no disability 164 54.7
Moderate disability 43 14.3
Severe disability 93 31.0
TOTALS 300 100.0
House-Confinement!
(During 2 weeks before hospitalization)
Outside on 3 or more days 159 53.0
Outside on 1 or 2 days 43 14.3
Confined to house 98 32.7
TOTALS 300 100.0

 

 

 

1See definitions in Chapters 5 and 13.

five weeks on the average, during which rehabili-
tation was a major goal. A comparison of
disability in this group with that in the general
U. S. population may be obtained by comparing
the data with the results of household interviews
in the National Health Survey of the civilian,
noninstitutional population in the United States
(10). Between the ages of 45 and 64, 64 per cent
reported that they had one or more “chronic
conditions”, and 20 per cent stated that they
had some limitation in major activities such as
working or keeping house. Over the age of 64,
these figures rose to 81 and 49 per cent,
respectively. In comparison, one-half to three-
quarters of our sample needed assistance with

personal care or walking. Since the inability to
walk or to care for personal needs implies
greater degrees of dependency than does inabil-
ity to work or keep house, it appears that most
of the members of this study are more severely
disabled than their contemporaries. For many,
the goals of treatment must be sharply limited.
Often, one thinks in such terms as eliminating
the need for a night attendant or enabling the
patient to manage by himself for a few hours
rather than in terms of full physical independ-
ence and return to work.

Socioeconomic Status and Function

Patients referred to Abington House, which is a

21

 
Effects of Continued Care

voluntary institution without direct access to
public funds (at the time of the study), are more
likely to be those who have some personal
financial resources. Patients in the same geo-
graphic area who are dependent upon public
assistance are usually referred to a large chronic
disease and rehabilitation hospital operated by
the county. The subjects of the study, therefore,
had generally more economic resources than
those of a related study, who had been dis-
charged from a municipal chronic disease hospi-
tal (8).

About one-third of these elderly, disabled
people were married and living with a spouse,
while more than a third were widows or widow-
ers (widows predominating over widowers by
more than 5 to 1). The remainder were divorced,
separated, or never married (Table 3.4). The

presence of a spouse in the home is clearly an
important resource when a patient needs per-
sonal assistance. Others, particularly children,
may substitute for a spouse. The factor of
household composition is, therefore, an impor-
tant one to consider in interpreting results.

The sample shows a broad distribution in
terms of social class, with a few more in the
upper two classes according to the Hollingshead
scale of social class than might be predicted
from studies of the general population (Ta-
ble 3.4). In a sample of households in New
Haven, for example, Hollingshead found 12.4
per cent in Classes I and II, compared with 19
per cent here (11). Among the present subjects,
12 per cent were college graduates, and another
31 per cent had completed high school.

The “Index of Economic Dependence”,

Table 3.4—Socioeconomic Status

 

 

 

No. of Per cent
persons of total
Marital Status
Married, living with spouse 111 37.0
Widow or widower 119 39.7
Divorced 24 8.0
Separated 3 1.0
Never married 43 14.3
TOTALS 300 100.0
Social Class by Hollingshead Scale!
I Professionals and executives 13 4.3
IT Lesser professionals and executives 44 14.7
III Small businessmen, administrative assistants, clerical
and white collar workers 64 21.3
IV Manual, skilled production workers 112 37.3
V  Unskilled and semiskilled workers 67 22.3
TOTALS 300 99.9
Index of Economic Dependence!
1. Independent 22 7.3
2. Partially independent 111 37.0
3. Partially dependent 155 51.7
4. Dependent 12 4.0
TOTALS 300 ~ 100.0

 

 

 

 

 

 

 

 

 

1See definitions in Chapters 5 and 13.

22
which we describe in Chapter 5, is based (for the
intake evaluation) on work status, home own-
ership, and source of economic support. Seven
per cent were independent, meaning that they
owned their own homes, were employed, and
did not receive support from a charitable agency
(Table 3.4). Most of the subjects, as might be
expected from their age and disability, were not
fully independent in these terms, although only
four per cent were completely dependent in that
they were not homeowners, were not employed,
and were receiving agency support (exclusive of
social security).

Examination of the economic factors in more
detail showed that a fifth of the patients had
worked up until the onset of the illness which
led to this hospital admission, while two-fifths
had retired within the preceding ten years.
Considering the men only, 21 per cent had not
worked for ten years or more, while the same
was true of 58 per cent of the women. As has
been demonstrated in other studies of the aged,
women continued to function as homemakers
longer than their spouses continued to work.
Over half the individuals in the study (54 per
cent) were active homemakers until the onset of
their current illnesses. This figure included thir-
teen men.

One indication of the relative affluence of the

subjects of this study is the fact that 63 per cent
owned their own homes (or their spouse did). A

majority (60 per cent) were receiving social
security payments. Private sources of personal
income, including pensions, annuities, and in-
vestments, were available in some measure to 83
per cent. Only 23 per cent received direct
financial support from members of their fami-
lies, and only 4 per cent from public or
voluntary charitable agencies.

Social Interaction

The social environment of a person who is
elderly, disabled, or chronically ill has a great
deal to do with how his needs are met. The
people in his household are potential providers
of service. The degree to which he interacts with
them and with more remote friends and relatives
influences their availability as resources. The
patient’s interaction with his social environment

The Patient Sample

may also be altered by a treatment program.
Because of the importance of such factors in the
course of a continuing illness, we took particular
care to define them and to describe them in
objective terms at the beginning of the illness.
Household composition, personal interaction,
and regular social contacts and activities are
summarized in Table 3.5.

In general, the subjects were integrated and
interacting with their social environment. For
comparison, in a study of 500 people over the
age of 60 living on the East Side of Manhattan
Island, 40 per cent were found to be living
alone, whereas this was true of only 28 per cent
of our sample (12). This proportion is compar-
able to the 25 per cent living alone found among
people 60 and older in a working-class borough
of East London (13). Two-thirds of the individ-
uals in the study had daily contact with a spouse
or relative and talked to three or more people
daily. Like most older people, they had with-
drawn somewhat from organized social activ-
ities; so that more than half reported that they
did not regularly participate in churches, clubs,
or the like. In the London study just referred to,
13 per cent reported that they went to church at
least once a month, and 12 per cent belonged to
a club.

The most common form of social relationship
for these elderly, chronically ill individuals was
with a spouse or child. Thirty-seven per cent
lived with a spouse, and 28 per cent with a child
or children. While 21 per cent reported daily
contact with a nonrelative, only 8 per cent
actually lived in a household with a nonrelative.
This group as a whole exhibited a degree of
family solidarity. Two-thirds had daily contact
with a spouse, a child, or a relative. An
indication of the extent to which these individ-
uals maintained their social roles within the
family is the fact that 61 per cent were reported
to be the heads of the households in which they
lived.

A group of patients of particular interest in
this study were those who were relatively
socially isolated and hence out of contact with
ordinary sources of assistance. There were 28
per cent who lived alone, 20 per cent who
reported no daily contacts with any other

23
Effects of Continued Care

Table 3.5—Social Interaction

 

 

 

 

 

 

 

 

 

 

 

No. of Per cent
persons of total
Household Composition
Married couple living with others 46 15.3
Married couple living by themselves 65 21.7
Solo* living with others 102 34.0
Solo living alone 85 28.3
Not living in a household or unknown 2 0.7
TOTALS 300 100.0
Personal Interaction
(Daily contact with spouse, relatives, or friends)
All 3 categories 8 2.7
Two categories 63 21.0
One category 165 55.0
No daily contact with anyone 59 19.7
Unknown 5 1.7
TOTALS 300 100.1
Persons Talked to Per Day
10 or more 73 24.3
6-9 38 12.7
3-5 101 33.7
0-2 69 23.0
Unknown 19 6.3
TOTALS 300 100.0
Regular Social Activities
More than one 61 20.3
One 73 24.3
None 164 54.7
Unknown 2 0.7
TOTALS 300 100.0

 

 

 

*Solo means widowed, divorced, separated, or unmarried.

person, and 23 per cent who talked to two
persons or fewer each day. There is considerable
overlap among these three groups.

Social Deprivation

Townsend, in his study of old people living in
London, makes a distinction between social
isolation and what he terms “desolation”, which
is a feeling of loneliness, usually related to the
recent loss of some accustomed social relation-
ship (13). Individuals in such a condition may be

24

included among those identified above as being
socially isolated. The isolated, however, also
include those who, by habit or preference, have
never associated much with other people. Be-
cause the most objective means of identifying
the desolated is by recording recent losses, we
have used the term “deprivation”.

The three kinds of losses which we considered
in the study were retirement from work, cessa-
tion of homemaking, and death of spouse. For
death of spouse, we defined the period of recent

 
loss as 2 years. For retirement from work or
cessation of homemaking, the 2-year period
excluded the 30 days prior to admission to
Abington House, since many patients who had
worked or managed their homes within this
recent period expected to return to their former
activities and did not perceive the change as a
loss, especially those who worked until the time
of admission to the hospital. On the basis of the
above definitions, 37.3 per cent worked or acted
as homemakers during the 30-day period before
admission to Abington House; and 30 per cent
had never functioned in these roles or had
withdrawn from them more than 2 years before
admission. The remaining 32.7 per cent were
defined as sustaining a recent loss in terms of
these social roles. Slightly more than one-third
of the patients (35.3 per cent) had experienced
at least one of the three social losses considered
here. Although these specific recent deprivations
did not affect a majority of the subjects,
three-quarters (222 patients) indicated that they
had experienced a constriction of their social
world, since they said that they talked to fewer
people every day than at age forty-five. This
finding may be compared with Cumming and
Henry’s report that two-thirds of a group of
middle-aged and elderly people in Kansas City
gave this answer to the same question, which
they define as a measure of “perceived life-
space” (14). Whether from a sociological point
of view we consider these experiences as a
diminution or a disengagement, they clearly
reduce the number of available social roles and
imply fewer regular social relationships which
could serve as sources of assistance.

Psychological Characteristics

The psychological and mental function of a
patient who enters a program for treatment of a
‘disability may be one of the major determinants
of his potential for improvement of deteriora-
tion. Although a public health nursing program
is not directed primarily at the improvement of
mental status, it does include treatment goals
having to do with the patient’s reaction to his
illness, interactions among the members of his
family, and his general psychosocial adjustment.
Thus, we may look upon the psychological and

The Patient Sample

mental functions of the patients in this study as
characteristics which limit or enhance the possi-
bility of improvement and, to a lesser degree, as

measures of outcome.
Many measures of intellectual and psychoso-

cial function have been proposed, but few have
been standardized for older populations (15).
Tests such as the Wechsler Adult Intelligence
Scale which have been so standardized are
lengthy and not well adapted to use when much
additional data must be collected (16,17). We
chose the following three measures which are
described in Chapter 5: (1) a test of “orientation
and mental control”, consisting of standard
questions such as “what is the date?” (2) a test
of ‘“‘observation and clear thinking” (Raven
Progressive Matrices), consisting of sets of pat-
terns to be matched, and (3) a rating of
“psychosocial adjustment” made according to a
standard distribution (Q-Sort) method. This last
rating was done by observer members of the
research staff at intake and was checked by an
independent rating made by one of the physi-
cians.

We summarize the results of these measure-
ments in Table 3.6. They may be compared with
a study of 208 patients (average age 60) dis-
charged from the county chronic disease hos-
pital referred to previously, in which these same
measures were used (18). The patients in our
study were oriented and functioning relatively
well mentally, since three-quarters were classi-
fied as ‘“‘clear”’, meaning that they made no more
than four errors out of a possible eighteen on
the test of “orientation and mental control”. By
contrast, only 52 per cent of the county hospital
patients were found to be “clear”. Only a small
number in both studies were badly confused
(2.0 and 8.0 per cent, respectively), while 16
and 22 per cent were partially confused in the
two studies. A much higher proportion of
patients were reported as untestable or untested
in the county hospital series (18 per cent).

The test of “observation and clear thinking”
showed that our subjects obtained scores in the
upper and lower thirds of the range more
frequently than did those discharged from the
county hospital. In the latter study, how-
ever, about twice as many were found to be

25
Effects of Continued Care

Table 3.6—Psychological Characteristics

 

 

 

 

 

 

 

 

No. of Per cent
persons of total
Orientation and Mental Control!
Clear (14-18) 235 78.3
Partially confused (5-13) 49 16.3
Confused (0-4) 6 2.0
Unknown 10 3.3
TOTALS 300 99.9
Observation and Clear Thinking (Raven)!
Good (28-36) 62 20.7
Average (9-27) 181 60.3
Poor (0-8) 18 6.0
Unknown 39 13.0
TOTALS 300 100.0
Psychosocial Adjustment Rating!
Good (16-40) 154 51.3
Average (42-66) 78 26.0
Poor (68-92) 38 12.7
Unknown 30 10.0
TOTALS 300 100.0

 

 

 

Igee definitions in Chapters 5 and 13.

untestable (‘unknown’). The median score for
our subjects was 20.7, which is very close to the
median score of 20 for “normally healthy old
people” (average age 75 years) reported by the
authors of the test (19).

The ratings of “psychosocial adjustment”
offer difficulties in interpretation. Comparison
of the ratings made of the same patients at
intake by physicians and observers showed
certain disparities, with indentifiable biases to-
ward higher or lower ratings associated with
different raters. The ratings in Table 3.6 are
those made by the nonphysician observers. The
range of possible adjustment scores is presented
with the upper and lower fourths separated from
the middle half. On this basis, the subjects of
our study, compared with those in the county
hospital study, showed roughly similar propor-
tions in the good, average, and poor adjustment
groups.

In summary, most of these elderly, disabled
individuals were able to function on a mental

26

and psychosocial level not greatly different from
others of their age and probably somewhat
better than the average for patients discharged
from a county chronic disease hospital. Rela-
tively small groups of persons had distinct
limitations in mental and psychosocial func-
tions.
References

1. Testoff, A. and Levine, E.: Nursing Care Supplied to
Older People in Their Homes. Am. J. Pub. Health
55: 541-547, 1965.

2. Associated Hospital Service of New York: Home
Care Following Hospitalization: Permanent program
of post-hospital care, 1960 Experience. Published
by the Associated Hospital Service of New York.,
N.Y, N.Y, pp. 1-30, 1965.

3. Phillips, H. T.: The intermediate
N.E. J. M. 276: 1352-1354, 1967.

4. Goldman, F.: Patients in chronic disease hospitals:
A Profile. A. J. P. H. 52: 646-655, 1962.

5. Annual Report for 1966: The Visiting Nurse
Association of Cleveland, Ohio.

hospital.

 
6.

U. S. National Center for Health Statistics: Persons
Receiving Care at Home, July 1958-June 1959.
U.S. Public Health Service, publication No.
584-B28, Series B, No. 28, U.S. Dept. of Health,
Education, and Welfare, Washington, D. C., 1961.

. U.S. Book of Facts, Statistics and Information,

1967: Published by Washington Square Press, Inc.,
N.Y, N.Y, p.22, 1966.

. Posman, H., Kogan, L.S., LeMat, A. F. and Dahlin,

B.: Continuity in Care for Impaired Older Persons:
Public Health Nursing in a Geriatric Rehabilitation
Maintenance Program. Report prepared for the
Committee on Health by Harry Posman,
Department of Public Affairs, Community Service
Society of New York, 105 East 22 St., N.Y.,N. Y.,
pp. 126-148, 1964.

. U. S. National Center for Health Statistics: Vital

Statistics of the United States, 1965. Vol. II:
Mortality, Part A. Published by Dept. of Health,

Education, and Welfare, Public Health Service, Wash-

ington, D.C. 1967.

10. U. S. National Center for Health Statistics: Chronic

11.

Conditions and Activity Limitation, July 1961-June
1963. U. S. Public Health Service, publication No.
1000, Series 10, No. 17, U. S. Dept. of Health,
Education, and Welfare, Washington, D. C., 1965.

Hollingshead, A. B. and Redlich, F. C.: Social Class

12.

13.

14.

15.

16.

17.

18.

19.

The Patient Sample

and Mental Illness. Published by John Wiley & Sons,
Inc, N.Y, N.Y. pp. 66-136, 1958.

Kutner, B., Fanshel, D., Togo, A. M. and Langner,
T. S.: Five Hundred Over Sixty: A Community
Survey on Aging. Published by The Russel Sage
Foundation, N. Y.,N. Y., pp. 30-42, 1956.

Townsend, P.: The Family Life of Old People.
Published by Routledge and Kegan Paul, London,
England, pp. 166-182, 1957.

Cumming, E. and Henry, W. E.: Growing Old: The
Process of Disengagement. Published by Basic
Books, Inc., N. Y.,N.Y., pp. 96-105, 1961.

Fisher, J. and Pierce, R. C.: Dimensions of
Intellectual ~~ Functioning in the Aged. J.
Gerontology 22: 166-173, April 1967.

Doppelt, J. E. and Wallace, W. L.: Standardization
of the Wechsler Adult Intelligence Scale for Older
Persons. J. Abnorm. Psychol. 51: 312-330, 1955.

Eisdorfer, C. and Cohen, L. D.: The Generality of
the WAIS Standardization for the Aged: a regional
comparison. J. Abnorm. Psychol. 62: 520-527,
1961.

Stroud, M. W. Ill: Highland View Hospital
Follow-up Study, Personal Communication.

Raven, J. C.:. Guide to Using the Coloured

Progressive Matrices. Published by H. K. Lewis and
Co. Ltd., London, England, 1960.

27
CHAPTER IV. PROVIDING CONTINUED CARE

Home nursing care for patients in the treatment
group was given by the Visiting Nurse
Association of Cleveland (VNA). This is one of
about 1800 agencies certified by the United
States Public Health Service to provide home
health services under Title XVIII, Public Law
89-97. Of the 1800, slightly more than thirty
per cent are visiting nurse agencies. In this
chapter, we describe the Cleveland agency and,
then, discuss nursing visits in the research
program.

The Visiting Nurse Association (VNA)

The VNA is a voluntary public health nursing
agency which has offered services in the Greater
Cleveland area for more than sixty years.
Initially organized to provide care for the sick in
their homes, its functions were soon expanded
to include prevention of illness and promotion
of health. The agency serves a population of a
little over one million for whom services are also
provided by six health departments and
numerous boards of education. The caseload
includes patients from a wide variety of social,
cultural and economic groups. Financial support
comes from the United Appeal, endowments,
other voluntary health agencies, and fees.

The VNA Board of Trustees, a citizen group
of forty members, has full responsibility for the
management and control of the Association.
They have directed the affairs of the agency
with a keen awareness of new developments in
the health field, accepted and promoted changes
in the interest of progress, and undertaken new
responsibilities with enthusiasm tempered with
thoughtful study. Further community participa-
tion comes through a public relations committee
with representation from various geographic
areas, racial, and ethnic groups. This committee
is responsible for interpreting the service, for
obtaining reaction of the community to the

28

service, and for offering recommendations for
strengthening the program.

In 1966, when the number of study patients
receiving service was at its peak, the staff
included:

Nurses:
Nursing Administrative Staff:
Director and Assistant Director 2
Supervisors 4
Home Care Coordinator 1
Staff Nurses:
Fully qualified (with college preparation) 41
Diploma graduates 12
Licensed Practical Nurses 5
Consultants:

Medical (part time)

Psychiatric (part time)

Physical Therapists (2 part time)
Occupational Therapist (part time)
Social Work (part time)

Nutrition (part time)

—— a = —

Comptroller 1

Secretaries and clerks 13

Responsibilities are assigned to nurses accord-
ing to levels of preparation and demonstrated
ability. A continuous effort is made to enable

the qualified nurse to devote a maximum
amount of time to those patients and families for

whom, in the judgment of the nurse and
supervisor, comprehensive nursing care holds
potential value. Various methods are employed
for using, in a team relationship, the qualified
public health nurse, graduate of a diploma
nursing school, licensed practical nurse, and
home health aide.

The VNA staff development program is
designed to help each nurse progress on an
individual basis according to her ability. It is
based on the philosophy that people learn and
develop best in an atmosphere which is
conducive to creative thinking and freedom of
expression and which provides opportunity for
nurses to participate in planning and policy
making. An organized staff education program
includes group and individual conferences.
Supervisors are required to have a master’s
degree from a program providing work in
administration and supervision. They devote
approximately fifty per cent of their time to
individual conferences with nurses. Physical
therapists function in a consultant capacity,
assisting nurses with the physical rehabilitation
aspects of their work. They visit for demonstra-
tion and supervision of exercises that are carried
out by nurse, patient, or family. Social workers,
nutritionists, and two physicians, one a psychi-
atrist, also provide consultant services.

The agency fee is based on average cost per
visit with adjustments made when the family
cannot meet the full payment. Since our intent
was to have the process in the research program
conform as much as possible to ongoing
practice, we decided to charge the usual fee.
However, to eliminate the possibility of having
cost influence acceptance of service, we advised
the nurses to waive fees for study patients when
in doubt about family willingness to pay for
services. Fifty-three or approximately one-third
of the patients in the care group paid a fee. Four
of the fifty-three paid full fee for all visits.
Twelve paid full fee for part of the service, and
thirty-seven paid partial fee for all or some visits.
Determination of the patient’s fee status was an
integral part of evaluation and planning for care
and often not established until several visits had
been made.

Care and rehabilitation of the sick is a
primary function of the VNA and is given with
an ever-widening perspective of the nurse’s
responsibility as therapist, health counselor, and
coordinator. In recent years, the agency has
intensified efforts to make available related
services, including those of the homemaker and
home health aide, occupational therapist, speech
therapist, dentist, and caseworker. Few, if any,
patients receive as much service as they need
because of limitations in financial support and
qualified personnel. The amount of service given

Providing Continued Care

cach patient is determined with consideration
for its potential value, the complexity of the
problem, interest in care as demonstrated by
patient and family, and the importance of
nursing and related health services to the total
treatment plan. When giving care to patients in
their home environment, it is essential that
consideration be given to the family as a unit
and with focus on the problems which are of
major concern and interest to them. Nursing
care encompasses psychological, social, cultural,
and economic factors. Coordinated planning,
when a multiplicity of professional personnel are
involved in the care, not only makes possible
common goals and greater economy in the use
of professional time, but also results in sounder
judgments on the part of each person who is
contributing to the care (1).

We specified in the research design that public
health nursing care would be given in the usual
service setting with four special provisions:

1) Nurses would be selected to participate in
the care program from those who had
demonstrated ability to provide compre-
hensive care and expressed interest in
taking part. In the interest of continuity of
care, priority would also be given to those
who expected to remain in the employ of
the agency.

2) All patients would be followed for 24
months and would receive visits as needed
with a minimum of one every three weeks
for the first 12 months and one every six
weeks thereafter.

3)A special record system designed to
describe nursing service would supplement
the agency case records. Activity records
would be completed at the conclusion of
each visit, and service given described under
broad categories of function. Narrative
reports would give information about
patient progress, goals in care plans,
problems, social and physical environment,
and contacts with physician and other
members of the care team.

4) A public health nurse coordinator would be
employed to serve as liaison between the
investigators and the agency personnel.

Continuity of care and minimal changing of

29
Effects of Continued Care

nurses is the usual practice in the VNA, but
particular effort was made to maintain this
standard for our research patients. Of the 150
patients visited, 100 were assigned to the same
nurse throughout their care; 38 had two nurses;
11 had three; and one had four. Changes in
nurses were explained by resignations of nurses
and changes in residence of patients when
distance made it completely impractical for the
same nurse to continue care.

Although care was given by only a small
proportion of the staff, participation in the
study was a total agency project. The Board of
Trustees, in the spring of 1962, paved the way
by devoting a Sixtieth Anniversary luncheon
program to the role of the agency in research.
All staff nurses were kept abreast of plans, and
the VNA nursing administrative staff and nurse
coordinator for research took responsibility for
policies and helped design research record forms.
The nurse coordinator consulted with super-
visors and staff about special records, patient
visits, and referral procedures, while supervisors
maintained responsibility for staff guidance
related to patient management.

Since the VNA staff frequently participated
in special projects, we anticipated that the
experienced nurse would assume her additional
responsibilities for research patients without
special difficulty. However, the following
problems were not anticipated and were of real
concern to the nurses. Two days before
discharge, physicians on our research staff
explained reasons for referral to the nursing
program to patients and families, and this timing
occasionally restricted the amount of discharge
planning. A very few families had already
employed a practical nurse or attendant, while
some family members who had come to the
hospital for instructions felt capable of giving
care independently. The nurse on her first visit
sometimes found a family fearful that the VNA
referral reflected on their ability, and a few
patients in the study lacked conviction that
follow-up care was needed.

Although two nurses were added to the staff,
referral of research patients could not be spread
evenly, occasionally resulting in peak loads and
necessitating deferment of visits to nonresearch

30

patients. Nurses also had to make the adjust-
ment of conforming to our research design
which specified the minimum number of visits
they were to make during stated intervals, when
they had been accustomed to making independ-
ent decisions based on the needs and interest of
patients.

The special “activities record” required the
nurse to analyze each aspect of a single visit and
record her various activities and functions under
specific categories. Although recording is an
accepted public health nursing function, some
nurses found the special records difficult. Most
of the nurses, however, reached the conclusion
that the research records helped them to plan
care more systematically. Supervisors com-
mented on the value of the activity records in
helping nurses to analyze situations in a more
orderly manner. This value, the supervisors
believed, was carried over into other cases.

As we noted previously, the supervisors and
coordinator assisted the staff with the special
responsibilities for research patients. Group con-
ferences with other members of the research
staff helped the nurses gain a more thorough
knowledge and understanding of the research
goals, to feel a part of the total program, and to
overcome fears and conflicts. Nurses had the
opportunity to share their problems and ap-
proaches to patient care with emphasis upon
increased understanding of the aging, the chroni-
cally ill, and their families. We stressed the
importance of their nursing contribution, and
the nurses increased their appreciation for the
value of long term supervision of patients who
have a chronic illness—less often possible in their
regular caseload because of too few staff. They
gained a fuller understanding of the meaning and
components of rehabilitation for elderly people
in their home settings. They also believed that
the study offered them an opportunity to make
a contribution to their profession. The nurses’
statements referring to the study changed from
the “hospital’s research’ to “our research”.

Our decision to test the study hypothesis in
the usual service setting had the practical ad-
vantages of an organized service with established
procedures for interagency cooperation and
coordination, personnel meeting recommended
qualifications, administrative structure that
met recommended criteria, affiliation with
other community agencies, and an organized
staff education program directed by qualified
supervisors and consultants. Participation in the
study was of value to the agency. Though we did
not design the study to examine all public health
nursing contributions to patients with chronic
illness and to their families, an opportunity was
provided for systematic examination of a broad
range of potential contributions. This jeint
endeavor by physicians and nurses is of special
significance because of the urgent need for more
comprehensive understanding about the needs of
the sick in their homes. This understanding is
essential for sound community planning and
effective use of health personnel at a time when
demands for care are growing at a rate far in
excess of available resources.

Service by Visiting Nurses in the
Research Program

The major concern of our study was the
comparison of outcomes in the nursed group
with those in the non-nursed group. Accord-
ingly, it was necessary to be satisfied that those
in the nursed group actually did receive nursing
care and that those in the non-nursed group did
not, before accepting seriously any conclusions
about differences in outcome between these two
groups. We expected that some degree of nursing
care would be sought by the non-nursed group
since, as was mentioned earlier, about 14 per
cent of patients discharged from Abington
House had, in previous times, been referred to
the VNA. Occasionally, also, patients who were
discharged from the hospital sought nursing
assistance independently. Not all of those
assigned to the nursed group received nursing
care since a few patients were unwilling to
cooperate with nurses and since a few died
shortly after discharge from the hospital. If the
expectations of the research design were met, we
would have expected every one of the 150
patients in the nursed group to receive at least
one visiting nurse visit and none of the 150
patients in the non-nursed group to receive any
visit. A crude measure of the degree of
adherence we actually observed is given in Table

Providing Continued Care

4.1, which summarizes the overall two-year
experiences of the patients in this regard. The
information in the table indicates that nonadher-
ence in the nursed group was slight (5/150),
while nonadherence in the non-nursed group was
somewhat greater (34/150).

The 34 in the non-nursed group who received
nurse visits had, with little doubt, different
characteristics from the 116 who did not receive
visiting nurse visits. Their elimination from the
non-nursed group would, thus, constitute a
selection bias of unknown magnitude in any
comparison of outcome between the two groups
and would prevent any meaningful interpreta-
tion of results. On the other hand, if the original
randomization assignments are accepted, an
equivalent proportion (34/150) of spontaneous
referrals to the VNA would be expected in the
nursed group if this group did not receive
assigned nursing visits. This is true by virtue of
the randomization process and the laws of
probability which apply thereto. Since the two
originally randomized groups are comparable
with regard to the likelihood of spontaneous
referrals to the VNA, the original random
assignments to the experimental groups were
strictly followed in all comparisons of outcome
throughout this report.

The amount of nonadherence to the study
plan was not as large as might be inferred from
Table 4.1 since, in this table, a person receiving
only one visit is given the same weight as a

Table 4.1 Numbers of People Receiving Visiting
Nurse Visits During the Study

 

 

 

Experimental | At least one LL
. No nurse visit | Totals
Groups nurse visit
(number of persons)

Referred to
Visiting 145 5 150
Nurse (N+)
Not Referred
to Visiting 34 116 150
Nurse (N-)

Totals 179 121 300

 

 

 

 

 

 

31
Effects of Continued Care

person receiving 100 visits. In terms of total
numbers of visits during the study, the nursed
group received 5667 visits, and the non-nursed
group received 834 visits. The frequency
distribution which we present in Table 4.2
contrasts the total number of visits per person in
the nursed group with the total number of visits
in the non-nursed group. In the nursed group,
the most frequent range of total visits was 21 to
40 visits per person, while the most frequent
range was 1 to 5 visits per person in the
non-nursed group.

We consider, next, the extent to which
visiting nurse follow-up was sustained. There
were 89 persons in the nursed group still
available for nursing visits (i.e., alive and not in a
nursing home) at the end of the study, and there
were 91 such persons in the non-nursed group.
In Table 4.3, we summarize nursing visits among
those available for visits at the end of the study.
Information in the table reveals that, at the end
of the study, 64 persons in the nursed group
were still receiving visits, while only 2 were
receiving visits in the non-nursed group.

From studying Tables 4.1, 4.2, and 4.3, we
conclude that (1) the number of people in the
nursed group who received nurse visits was much
greater than the number in the non-nursed

Table 4.2—Distribution of Patients According to
Number of Nurse Visits During the Study

 

 

Table 4.3—Numbers of Persons Receiving Visiting
Nurse Visits Among Those Available for Visits

 

 

 

at the End of the Study
E iment LL .
xperimental Receiving Not receiving | ois
groups Visits visits
(number of persons)
Referred to
Visiting 64 25 89
Nurse (N+)
Not Referred
to Visiting 2 89 91
Nurse (N-)
Totals 66 114 180

 

 

 

 

 

 

 

 

Experimental Groups
Number of nurse

visits during Referred to Not referred
the study Visiting to Visiting
Nurse (N+) Nurse (N-)

(range) (number of persons)

0 5 116

1-5 21 11

6-10 10 8

11-20 27 3

21-40 36 6

41-60 23 1

61-80 9 2

81-100 10 1

101+ 9 2

TOTALS 150 150

 

 

 

 

 

32

group; (2) of those receiving visits, the people in
the nursed group received more visits per person
than those in the non-nursed group; and (3) the
nursing care follow-up was more sustained in the
nursed group than in the non-nursed group.

The fact that there were large differences in
nurse visits between the nursed and non-nursed
groups is further strikingly demonstrated by
examining Figures IV-1 through IV-6. In these
Figures, nurse visits are presented graphically for
each patient of a 2-year horizontal time axis.
Breaks in the horizontal time bar indicate
intervals of time when patients were not
available for visits due to death, hospital stay,
nursing home stay, move out of the VNA service
area, or withdrawal from service by mutual
consent between the patient, his family and the
nurse. Figures IV-1 through IV-3 present the
graphs of nurse visits for patients in the nursed
groups, and Figures IV4 and IV-6 present the
graphs for the non-nursed group. We classified
patterns of visits of the graphs as high, medium,
or low frequency patterns, where the frequency
of visits was defined as the number of visits in
the 2-year study period divided by the number
of days available for visits in the 2 years. A
frequency of at least 0.10 was called “high” and
corresponded to at least one visit every 10 days,
on the average. A frequency between 0.10 and
0.05 was called “medium”. A frequency of less
than 0.05 was called “low” and corresponded to
less than one visit every 20 days, on the average.
£€

PATIENT
NO. SEX AGE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

7 F 80 [MM ITITIITITT BT TO CII TI TOIT TTI TT 1171
20 F 69 [III OT I I I TT TT 1)
23 F 72 [I ]
24 F 78 [III ]
28 F 70 IIIT TTT ]
42 F e7 [OMI ]
45 M 7! [IIT OC TOIT OO ]
46 M 55 [III  JTIMOMOMOOOOOOT TTT 1 [ [ 0
49 Mm 79 IH ]
57 M 74 [IIIT IO OOO OOO I Td CTT
59 M 75 [I I I I I CT I OI TTI TT TT TT 1]
60 F 71 [III IIIT IIIT TO IO OO TIC I IT TOIT 1
65 F 80 [IIIT TTT TIT TTT T CC T7171
66 F 61 [III III N ]
73 M 66 [MMT IIIT TT CIO OT TO TO TOO I TIO TIT 110
82 M 51 [IIL ]
84 F 79 [ MINI I TO TO OO I IT TI)
87 F S57 (_ ITI INST INT IIIT Jr ]
Nn F 72 [IIIT JOO OOOO TOO I TOOT IT ]
120 M 59 [I TC I OT TOT TT TT 11 ]
127 M 81 [ ae Ce re rr WT JT rrr rrr rrr rr mrirt 1
128 F 53 [IIIT ITT OOONOT OMT Ir mT 1 ]
129 M 60 MIMI TT TTI IIIT TIT TIT TT TOIT TT Ts ]
134 F 72 [IIIT OT TOTTI TTI [TTT TC IOI TTT TID
143 F 84 [ITI IMT TIT TOOT I TOT TO TOIT ]

 

IEach vertical line represents a nurse visit along the 2-year horizontal time axis. Breaks in the time bar indicate unavailability for visits because of hospital stay, nursing home stay,
move out of VNA area, refusal, or death. “High frequency of visits’ was defined as at least one visit per 10 available days (on the average).

Figure IV-1. Patients in Treatment Group with “High Frequency of Visiting Nurse Visits”!

are) panunuo)) Suipiaoig
x PATIENT
NO. SEX AGE

144 F 75 LUI

v

LE ET RTA

   

 

 

 

   

 

   
   

 

 

 

 

 

 

 

 

 

   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

146 F 86 [I JIITITITTITIIIITII J
156 M 69 [ JIIIIIIMIMIIIT J
57 M 7o [MMIII IT I TIT TI TTL TTT II ]
163 M 87 [IIIT III 00 TTT TL 11 I I A 1]
175 M 71 [ICT THT mi
176 F 53 [I ITI Eg AE TT

177 F se [IIH TEA TI nn ree
180 F 66 [IIIT IIIT TO mm ]
201 F 65 [IIIT ]
209 M Se [IIT IIIT IIIT ]
217 F 75 [MMMM TTTTTT TOT TIT TIT TOT IT TO TT CJ
29 M 72 [I J
226 F 68 [IMT TT TITITIIT ]
227 M 75 [IIIMIMIMII TIT TIT IIL 10 ]
229 M 56 [TI I I TI CI TI TT TT ]
231 F 71 [III TIT TOOT. TIO ]
250 F 55 (ITI TINT ]
251 F 7s [IIIT ]
28] F 94 (IIIT OT IIT TIT II1I J
289M 75 IIIT MMI II ITT TTT TTT ]
290 M 77 [IMM TTI TTT TMI TOOT TT TIT TT TTT TT T1T1D0
297 M 75 [CLIT OOO OC OL TTT [IIIT 0
302 M 68 [MII TIT TIITITT 1

IEach vertical line represents a nurse visit along the 2-year horizontal time axis. Breaks in the time bar indicate unavailability for visits because of hospital stay, nursing home stay,
move out of VNA area, refusal, or death. “High frequency of visits” was defined as at least one visit per 10 available days (on the average).

Figure IV-1 (contd). Patients in Treatment Group with “High Frequency of Visiting Nurse Visits”!

aIe) Panuuo)) Jo s303JJ4
SE

PATIENT

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

NO. SEX AGE

15 M 60 [TTI ]
22 M 73 IIO TIT 13I ) ]
29 F 68 (MMMM TT TTT TT 1 TT Tor TT ]
33 fF 73 OMIT TO TMI TIO TIT OT TT TT TTT TTT 011] C11 1]
35 F 78 [IIMMMMIITOTIT T [TT TT TT TT TT T1 I I 1
36 F 72 IIIT LT IT IT IT T1171 ]
4] F 90 MINT ITI MMITITITIT THI TOT TIT T TTT 11
43 F 69 [TT T71 ]
52 F 79 [IIIT Toor TT ]
53 F el CITT 1 ]
64 M 71 [IITITTII [| I TMT
69 F 85 [IMT IO IT TTT ]
70 F vo ITT T TTT NMONOTT TT TT T TTT TTT ]
88 F 83 [MIIIITIT TITITIIT LTT 1 | 1 1 1 71 1 11 11
91 F 73 [III [TT 1 T [110 JI ]
92 F 51 OIC ]
97 F 7 MIIII'TTT TIT TIITT TT ITT TIT ITTITT TT TTT TT 71 111
0 F 84 III IT IT IT TT ITT TT TT 1 TT TT TT TT 1]
02 F 79 (IIIT TI” IIT T TTT TT TTT TTT TT TTT TTT
105 F 65 MII] EINE — IC 1 I THT
06 F 84 [ MITT IIIT III ITO TIT TIT TITTT TTT TT T7117
mo fF ee MITT TIT TIT TT TTT T TTT 1 ]
8 M 59 MINI —TT TTT] ]
131 F 89 [IIIT 1111 J
4 F 74 [ TTTTT TT 1 ]
145 M 77 [CITT CL 11 I CT TOO TOOT IT TT IIT TTD
147 F S52 (MIT IIT TT OT TT ION TM TIT TTT TTT T1171

 

Each vertical line represents a nurse visit along the 2-year horizontal time axis. Breaks in the time bar indicate unavailability for visits because of hospital stay, nursing home stay,
move out of VNA area, refusal, or death. “Medium frequency of visits” was defined as one visit per 11 through 20 available days (on the average).

Figure IV-2. Patients in Treatment Group with “Medium Frequency of Visiting Nurse Visits”!

aIe)) panunuo)) Surpiaoig
& PATIENT

9

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

NO. SEX AGE

50 F 7 [TIT OTIT ITT ITT TIT IIT TTT TT TT10 1]
158 F 86 [I II ]
167 F 80 IIIT IMT IIIT TIT ION ITI TIT TIT T° 1 1 1 ]
179 F 78 (I [TT 1 TTI TTI [TT TOTTI TIT TTIT
183 F 88 [II ]
190 F 54 [MITT TINT T T_T ITTTITITITTIT 1 FTI  rrfrr
193 F 75 [ IIITI MMT ITT TIT IT TT TITTT J
194 M 72 [TOC TOOT TOOT TT TTT TTT 17 11 1 ]
195 F 69 WL fr | L111] 7]
97 F 74 IL IIIT TT T1 J
199 F 70 ul [11 [IIIT I LITE III ]
200 F e6 [MIL IT ITTTTIT TTTTIITIT 1 ]
213 M 73 LL LOH gr | 1 | ]
24a Mm 78 [III TT — IT IT TIT TIT 1 Il ]
233 F 69 [II IITTTTITHrI1 J 1 [I TT IT I'TT1T1 TT TT TTT 1-1 ]
237 F 89 (IIIT IT IT TMT IT TT TTT ]
238 F 83 [IIIT ITT T TTT TTTT] ]
242 M 56 (IIT ITT TIT TIT ITIIIT | [1 | [1 | | 1]
253 F 78 [MII ]
259 F 66 [IL IITT IMIITTT TIT 1 [TT TTT [TIT TIT TTT IT IT IT
267 F 74 [IL I A I [TIT T TT1T11 JT OWT ITTITITTD
279 F 72 (IIIT MMIII mmr Ir momo TTT TT 171 0
283 M 59 [ITT ITTIIM TT [1 TOOT Orr rm [MMT ITI rT
285 F 52 [IT] ]
291 M 64 (IMM TMT TIT MIT ITT TIT T1177] ]
293 F 50 [IIIT TIT 1) - ]

 

IEach vertical line represents a nurse visit along the 2-year horizontal time axis. Breaks in the time bar indicate unavailability for visits because of hospital stay, nursing home stay,
move out of VNA area, refusal, or death. “Medium frequency of visits” was defined as one visit per 11 through 20 available days (on the average).

Figure IV-2 (cont’d). Patients in Treatment Group with “Medium Frequency of Visiting Nurse Visits”!

aIe) Panunuo)) Jo $399J5q
LE

PATIENT

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

NO. SEX AGE

IF 57 [ CL TIO IIIT TTT TT TTT TTT 1
25 M 62 [TT 11 T_T TT TT 71 1 ]
37 F 77 LITT IT TT TT 1 71 1 T ]
54 F 79 MII IL I I I I I ITI TIT T TT TT TT TIT T1 ]
71 F 59 [IIIT 11 I ]
90 F 67 [IT 1 I I TT ITT T1 | I | 1 J
100 F 63 MI IIL I ITT T TI 1 1 I I ]
12 F el [LTT I [ I I 0
4 F 70 [TI 11 TT ]
nz fF 76 III I_ TL [11 1
119 F 64 [ILI T TI 1 I ]
149 M 84 [IT TT TT [ TIT TT [TTT ] ]
I51 F 85 [I] [T1111 T 1 ]
68 M 70 (MIMI TT TTT TT TT T7 1 CT 1 T]
169 M 80 [TT 1 TT °C ]
70 M 70 [LL IIT TT TT TT TT TT T7 T1 1
172 F 59 (III III TT TT T TT TT 7 7
178 F 73 [II__ TTT T [TT I I I [TT 7]
91 F 87 [LLL I [ T T T1 I I
207 M 80 [TC ]
210 F 78 [ITT T1 I IT T_T T 1]
21 M 80 [II ITT 1 I ]
230 F 59 [MI IT [TT TT TT 1 [ ]
254 M 68 [TT IITIT [IT [TTT 7 I I I I 7
257 M 91 [T_T IT ]
262 F 75 [TTT TTT TT TT TT TTT CT TIIT ]
266 M 83 [IL TI TT 1 I 1 0

 

Each vertical line represents a nurse visit along the 2-year horizontal time axis. Breaks in the time bar indicate unavailability for visits because of hospital stay, nursing home stay,

move out of VNA area, refusal, or death. “Low frequency of visits” was defined as less than one visit per 20 available days (on the average).

Figure IV-3. Patients in Treatment Group with “Low Frequency of Visiting Nurse Visits”!

are) panunuo)) uiprAolg
wo

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

PATIENT

NO. SEX AGE

273 M 60 [MII IT T1111 | I TE 1d | | | | 11]
274 F 75 (II 1 ]
275 F 84 [II] | [ _ | | ] | 1 1 i]
278 F 78 ILL I [ © TI 1 1 [ [ ]
280 M 66 [IIIT ITT TTT] OTT 7 7 1
294 M 81 [IIIT IT J
299 F e6 LI TIT T 1 [ J]
300M 51 CIIOOIIT TTT T TT 1 I I I

 

LEach vertical line represents a nurse visit along the 2-year horizontal time axis. Breaks in the time bar indicate unavailability for visits because of hospital stay, nursing home stay,

move out of VNA area, refusal, or death. “Low frequency of visits” was defined as less than one visit per 20 available days (on the average).

Figure IV-3 (cont’d). Patients in Treatment Group with “Low Frequency of Visiting Nurse Visits”!

QI) Ponunuo)) Jo $3093
6¢€

PATIENT

  

 

 

 

 

 

 

 

 

   

NO. SEX AGE

9 F 79 [ I ]
30 ™M 58 (TIMI IIT TIT ]
74 F 71 [ HI ]
76 F 66 [1 7
78 M 52 J
I09 F 67 RNETEY TTR COC LCV FO vECIVECTTOOTT Creer nove CETCrrurovyrrrory Bovey 1 Ome
1n3 ™M 70 ]
132 F 74 ]
184 F 83 [IIIT IIT — J
198 M 51 [III ]
234 M 79 [HITT ]
243 Mm 79 L I OT IT IIT TTT 10
247 F 62 [ IIMMMIMIIIIIOIIIOOIT —[ II ]
248 F 59 [  IITTIITTTTTHTMIMNEEMNINI TTT T7170 ]
249 m 80 [IIL 3
265 M 87 [ IINIRRIRRLRRRRRR RINNE IARI nnLniN ]
268 M 77 TT OI TC IT TT TIT TT TIT] 3
276 F 76 [III ]
301 M70 [Arar ]
Each vertical line represents a nurse visit along the 2-year horizontal time axis. Breaks in the time bar indicate unavailability for visits because of hospital stay, nursing home stay,

move out of VNA area, refusal, or death. “High frequency of visits” was defined as at least one visit per 10 available days (on the average).

Figure IV-4. Patients in Control Group with “High Frequency of Visiting Nurse Visits’!

are) panunuo)) Surpiaoig
5 PATIENT
NO. SEX AGE

0

 

67 ITT TT TT ITT1 ]
74 (TTI T1

64 [ MITT T TTT TT TT TTT T1
94 84 [IIIT

138 78 [IIIT]

IEach vertical line represents a nurse visit along the 2-year horizontal time axis. Breaks in the time bar indicate unavailability for visits because of hospital stay, nursing home stay,
move out of VNA area, refusal, or death. “Medium frequency of visits” was defined as one visit per 11 through 20 available days (on the average).

39
67
79

 

 

mM Mm
Jil J

Figure IV-5. Patients in Control Group with “Medium Frequency of Visiting Nurse Visits’!

aI) PAaNuUIIu0)) Jo $}O9JJ5
Iv

PATIENT
NO. SEX AGE

69 [ IMT 1

75 [ 1 I
76 [I

225 51 [I T

228 8a [ T mJ

IEach vertical line represents a nurse visit along the 2-year horizontal time axis. Breaks in the time bar indicate unavailability for visits because of hospital stay, nursing home stay,
move out of VNA area, refusal, or death. “Low frequency of visits” was defined as less than one visit per 20 available days (on the average).

SI

63
164

 

 

dl) dea

 

Mmmm g

Figure IV-6. Patients in Control Group with “Low Frequency of Visiting Nurse Visits]

are) ponunuo)) SUIpiAoIg
Effects of Continued Care

Patients who were never visited or who had
fewer than 10 available days for visits during the
study were considered ineligible for frequency
classification (13 in the treatment group and
121 in the control group); therefore, graphs for
such patients are not included in the figures. In
the nursed group, the frequency of nurse visits
over the entire 2-year period was high in 49
patients, medium in 53, and low in 35. In the
non-nursed group, the frequency of visits was
high, medium, and low in 19, 5, and 5 patients,
respectively. The remaining patients in each

1001
95
90
85
80
75
70
65
60
55F
S50
45
40r
35
30
25
20

 

CATEGORY PERCENTAGES

 

L A 1
° 3 6 9

group were either never visited or had fewer
than 10 available days for visits.

For patients in the group assigned to the
nursing program, we illustrate the decreasing
proportion of those receiving nursing services in
Figure IV-7. The decreasing proportion is
explained by withdrawals from the nursing
program by reason of death, admission to
nursing homes, and other reasons as graphed in
the same Figure. At the end of the study, 42.7
per cent of these patients were still receiving
nursing services: 12.7 per cent had withdrawn

O= WITHDRAWN FOR ENTRY
INTO NURSING HOME

0 = WITHDRAWN DUE TO DEATH

A = WITHDRAWN FOR OTHER
REASONS

X= NOT WITHDRAWN

 

1 1
12 1S 18 2 24

1 |

TIME SINCE ADMISSION INTO STUDY (MONTHS)

*The denominator for these percentages is the 150 patients who were assigned to the nursing program (N+ group)

Figure IV-7. Withdrawal from Nursing Program

42
because of entering nursing homes: 22 per cent
had withdrawn by reason of death: and 22.7 per
cent had withdrawn for other reasons. Among
the latter, 6 patients had moved out of the area
of VNA jurisdiction, and 5 had never received
nursing services.

Extent of Nurse Follow-ups (Examples)

 

Patient number 36 was hospitalized in
Abington House for rehabilitation after sustain-
ing a unilateral fracture of the tibial plateau. She
was 72 years old, had never been married, and
had retired about 9 years earlier from her
position as a buyer and seller of children’s books
for a locally well-known book and antique store.
She was relatively isolated socially and obtained
financial assistance from Social Security and
from a local private philanthropic social agency.
After discharge from Abington House, VNA
visits were initiated. A cleaning woman also
helped her at home. The patient complained
increasingly of gastrointestinal discomfort. She
became more isolated, expressed dissatisfaction
with the nurse, and was concerned about death.
After a period of about 7 months in which she
received 15 nurse visits, she had an episode of
hemoptysis and was rehospitalized. A gastric
resection was performed for carcinoma. Compli-
cations ensued, including intestinal obstruction
and myocardial infarction; and the patient died.

Patient number 87, a 57-year-old housewife
with rheumatoid arthritis, was assigned to VNA
assistance in the home after being discharged
from rehabilitative treatment in Abington
House. Though unable to do heavy housework,
she was independent in self-care except for some
© assistance in bathing. Her husband, also partially
disabled, could not assist; and the patient moved
from her own home to the home of a married
son. For 16 months, an active program of
nursing assistance was maintained, in which she
received 130 visits. The program was interrupted
only by five short-term hospitalizations. The
nursing visits supplied intensive assistance in
such areas as exercises, prescribed medication,
tub bathing, and psychosocial care. At the end
of 16 months, nurse visits were terminated since
the patient moved out of the state and, thus, out
of the VNA service area. The move was made to

Providing Continued Care

the home of an unmarried son, where her
husband also lived.

Patient number 92, a 51-year-old housewife,
had been admitted to Abington House for
rehabilitation after a stroke. Prior to the stroke,
she had been an active homemaker, taking care
of two adolescent children and participating in
ceramic, investment, and social clubs. Her
husband was independently wealthy and em-
ployed full time practical nurse assistance in the
home when the patient left the hospital. The
patient’s sister also helped, and the VNA nurse
attempted to establish a coordinated program of
assistance within the home. The patient was
aphasic and walked with the aid of a short leg
brace and quad cane. The family was unable to
adjust to her disability, and she was admitted to
a nursing home within six weeks after discharge
from Abington House. During the 6-week period
after discharge, she received four nurse visits.

Patient number 200 was a 66-year-old widow
with 2 married daughters. She had become
widowed 22 years earlier and had been
successful in both business and purposeful social
activities. Two years before admission to
Abington House, she had developed paraplegia
of unknown cause and had narrowed her

business activities to clerical work at home.
After discharge from Abington House, she

received VNA assistance (19 visits) for about 10
months in relation to skin care, catheter care,
and a prescribed program of exercise. A physical
therapist also came into the home on occasion,
and her daughters visited regularly to assist with
housekeeping. Though she made a determined
effort to maintain her strength and independ-
ence, the paralysis became noticeably worse
three months after she left the hospital. She
continued a downhill course physically, but
remained alert mentally. She had to move to the
home of a married daughter 10 months after
discharge. Since her daughter’s home was out of
the area of the VNA service area, VNA visits
were discontinued; and five months later, the
patient chose to enter a nursing home.

Patient number 52 was 79 years old and a
widow of 20 years duration. In Abington House,
she received treatment for an intertrochanteric
fracture of the right hip and for arteriosclerotic

43
Effects of Continued Care

heart disease with chronic congestive failure and
angina pectoris. A self-employed dressmaker
until age 69, her income was from past savings.
She had no children or relatives and had been
living a relatively isolated life due to age, illness,
and mild mental confusion. She lived in an
apartment and expressed fears about going out
alone. She received 18 nurse visits during a
10-month period after discharge from Abington
House and was also assisted by a home aide for
two or three days each week. The nurse devoted
much effort to supervision of medication and
attempted unsuccessfully to coordinate care for
the patient. Physician visits were spaced
irregularly, partly as a result of the patients
inability to hear telephones and doorbells. She
did not adjust well to her illness and voiced
many complaints. Because of dissatisfaction
with her physician, she went to a hospital
outpatient clinic. In the clinic, she also had
many complaints. She sought out another
private physician and complained about the
financial burden of private care. After 10
months at home, she suffered a heart attack, was
hospitalized, and died.

Patient number 226 was a 68-year-old lady
who had never been married and who had
retired from clerical work 9 years earlier. She
and an unmarried sister with whom she lived had
no outside interests and were very dependent on
each other. Financial support came from Social
Security and from a philanthropic social agency.
In Abington House, the patient was treated for
stroke, fracture of the wrist, arteriosclerotic

heart disease, and mild diabetes mellitus. She
was not paralyzed, but was demented and

incontinent. For about 10 and one-half months
after discharge from the hospital, she was visited
by the VNA nurse. The nurse, in 29 visits, dealt
primarily with problems of adjustment to illness,
assisting also with medications, bathing, and
exercise. These visits were interrupted by a
temporary period of care in a nursing home at a
time when her sister was ill. The patient had a
second stroke ten and one-half months after
discharge from Abington House. She was
rehospitalized for three weeks, remained semi-
comatose, and was discharged to care in a
nursing home. After one more hospitalization

44

for bronchopneumonia, she died in the nursing
home, having lived a total of about three months
after the occurrence of the second stroke.

k kok ok ck

We classified nursing activities for the study as
those concerned with exercises, prescribed
treatments, activities of daily living and
mobility, health promotion, psychosocial care,
and planning and coordination. In particular, the
activities which we recorded systematically were
those which reflected assistance to the patient,
the family, and, in the case of planning and
coordination, other service personnel. The
following are but a few examples of the many
combinations of services and activities which
were recorded. Certain patients were helped
with physical exercises while others were
instructed in exercises. The families of many
patients received instructions about exercises,
diet, or prescribed oral medication. Nurses
administered parenteral medication, performed
irrigations, and changed dressings. They assisted
patients with bathing and instructed families
about bathing procedures and about other
activities of daily living, including also ambula-
tion and stair-climbing. Health promotion
activities included instruction and assistance
with problems in such areas as nutrition,
accident prevention, hygiene, and recreation.
Nurses helped patients and families identify such
psychosocial problems as reactions to illness,
grief, interpersonal difficulties, and economic
problems. In the area of planning and
coordination, nurses communicated on behalf of
the patients with occupational therapists,
physical therapists, physicians, social workers,
and personnel in many other community
agencies.

We analyzed the amounts of nurse services
and changes in nurse services over time in terms
of the proportion of patients and/or families
involved in any given service from among those

eligible to receive nursing services within a
defined interval of time. On this basis, more
than 90 per cent of eligible patients in the
nursed group received nurse assistance in the
following areas during the first three months of
the study: activities of daily living (including
also mobility), health promotion, planning and
coordination, and psychosocial care. In the case
of health promotion and psychosocial care, the
proportion of patients receiving such assistance
remained greater than 90 per cent throughout all
the quarterly intervals of the two years of study.
In the case of activities of daily living (including
also mobility), the proportion receiving assist-
ance decreased slightly, but remained greater
than 80 per cent throughout the study. The
proportion receiving assistance in planning and
coordination decreased to 69 per cent in the
second three months of the study and varied
between 49 and 61 per cent thereafter. At the
beginning of the study, the proportions of
patients who were assisted with oral medica-
tions, prescribed treatments, and exercises were
75 per cent, 61 per cent, and 56 per cent,
respectively; and, by the end of the study, these
proportions dropped slightly to 59 per cent, 45
per cent, and 43 per cent, respectively.

As an example of a specific service within one

Providing Continued Care

of the broad service classes described above, 90
per cent of eligible patients in the nursed group
received ambulation assistance during the first 3
months of the study. The proportion remained
at this high level during the first year and then
decreased to 79 per cent by the end of the
study. As another example of a specific service,
85 per cent received nursing services in relation
to accident prevention at the beginning of the
study.

In summary, study patients referred to the
Visiting Nurse Association received continuing
care at home. These patients actually received
much more and extensive care than those not
referred to the VNA, as measured by the
number of people served, the number of visits
per person, and the degree of sustained
follow-up.

References
1. Reese, E.: Public Health Nursing and Compre-
hensive Health Care. Nursing Outlook 16: No. 1,
48-52, January 1968.

45
CHAPTER V.

Nurses, physicians, and sociologists were the
research observers. They were trained and
supervised by faculty with doctorate degrees and
advanced academic and applied experience in
biostatistics, internal medicine, preventive medi-
cine, nursing, psychology, and sociology. In
order to maintain the objectivity of observer
evaluations, we did not give them details about
the experiment’s design nor about the process of
random assignment of service and observation.
We, furthermore, kept them geographically apart
from service personnel to avoid communication
and exchange of information.

INTERVIEWING

Before the study began, we trained all
observers in the interviewing procedures and
gave them a guide which: 1) emphasized
principles that applied to this particular study;
(2) defined the content of the interview; (3)
defined the observations and measures of the
study; and (4) defined the sources of the
definitions and measures. During the periods of
observer training and data accumulation, the
following principles were stressed and applied as
consistently as possible (1).

The function of the observer was to establish
and maintain contact with patients and families,
to obtain accurate information without influ-
encing patient behavior, and to record this
information. In making first contact with the
patient in the hospital, the observer introduced
herself by name. She wore a white coat and,
though she introduced herself as a member of
the evaluating staff, she was occasionally
identified by patients as a member of the
hospital staff. She was attentive to questions
about her role and carefully defined her position
as nonservice. When the observer suspected that
the patient had some concern or negativism, she

46

INTERVIEWING AND MEASURES

tried to get the patient to verbalize this before
dealing with the problem. Explanations were
then given in response to expressed questions,
and anxieties related to unexpressed questions
were avoided. In order to get assistance from the
supervisory staff, the observer was given the
option to delay her answers. At the end of each
interview, she prepared the patient for the next
interview.

During the interview, the observer used simple
words which could be understood by the
patient. The patient or informant was encour-
aged to talk freely and comfortably. The
observer was attentive and ready to listen,
constantly aware of her role as one who gath-
ered information without influencing patient be-
havior. To this end, she did not give advice and
avoided answering questions about care. She
created a quiet, interested impression and was
conscious of the impressions which might result
from even such details as manner of dress. She
conducted an interview that appeared unstruc-
tured, although she had a structured plan of
interview content in mind, including also
specific questions, observations, and data forms.
She attempted to gain information in the course
of natural conversation, a task which was made
easier by the fact that the topics of research
importance were of personal concern to the
patient. The patient was, in fact, often inclined
to expand on the topics and to continue the
conversation.

The flow of conversation tended to follow a
predictable pattern that the observer directed in
subtle fashion, thereby gathering the required
data. Neutral topics opened the interview, such
as the weather, the observer’s experience in
finding her way to the patient’s home, and
objects of interest in the patient’s home. The
conversation was then directed to subjects of
research importance, generally concerned with
observable physical activities of the patient as,
for example, a discussion of the patient’s
walking status or of mechanical assistive devices
in evidence. Such remarks led easily into a
discussion of activities of daily living and the
nature of assistance received by the patient. The
topic of assistance was then followed by
discussions of household composition, family
structure, socialization, nature of medical care,
and management of economic affairs.

As a check on accuracy and reliability of
patient responses, the observer asked the patient
to show her such things as: (1) the bathroom
and (2) medications in another room or
meaningful substitute objects. These requests
created test situations for direct observation of
transfer, locomotion, and communication. They
served also as checks on the validity of
information about going to toilet, transferring,
and walking, and as indirect checks on the
reliability of interview information about
bathing and dressing, which require a degree of
mobility. Observersswere also able to form a
judgment about general reliability and made
other observational tests when indicated. It was
frequently appropriate to ask the patient to
demonstrate the location of the bathroom
during that part of the interview in which
bathing, dressing, and transfer were discussed.
The physical activity served as a break in the
conversation and gave the patient an oppor-
tunity to be active. In association with the term
“bathroom”, it was possible to discuss the more
personal areas of research interest such as
toileting and continence.

In the continuing conversation, such topics as
“future plans of the patient” and “people
interested in the patient” led to expanded
discussion of household composition, family
structure, social interaction, economic status,
and nature of medical care. Since the observer
tried to create the impression of an unstructured
interview, the patient’s conversation could easily
digress from the schedule of questions. The
observer continued to listen, keeping unan-
swered questions in mind, some of which the
patient answered in the course of this
open-ended conversation. Many questions, thus,
did not have to be asked. Rapport was

Interviewing and Measures

strengthened by the interested attitude of the
observer, and the patient felt more at ease than
would be possible under the circumstances of a
stream of directed questions. Since it was
necessary to influence the direction of the
course of conversation, interviewers were in-
structed in methods of changing topics directly
or indirectly and were provided with examples
of guiding statements.

Whether information was obtained indirectly
or by direct questioning, the interviewer was
trained to be sensitive to its accuracy. She was
aware that the patient was not familiar with
research definitions of even such commonly
used words as “bathing”, “dressing”, “friends”,
“employment”, and “doctor”. The interviewer,
therefore, used the patient’s own words and
referred to objects and events that were familiar
to the patient. The observer was instructed to
check the accuracy of the patient’s statements
against other sources or by means of rephrasing
questions. Many questions were often needed to
confirm the accuracy of a claim of independent
function, and confirmation often required
additional observations. Patients were inter-
viewed and observed wherever they lived. If the
patient was unable to give information, a
relative, friend, or attendant close to the patient
was interviewed. Unless otherwise stated, the
interval of evaluation was the 2-week period
which immediately preceded the due date of the
evaluation. Research forms were completed by
the interviewer immediately upon leaving the
patient.

Frequency and Completeness of Interviews

As described in the chapter entitled “The
Research Plan”, 150 patients were scheduled to
have additional interviews one week after
discharge from Abington House and every three
months after intake into the study. This group
was labeled ‘“‘the observed group”. The remain-
ing 150 patients did not have the additional
interval interviews and were termed ‘‘the
non-observed group”.

The process of carrying out interviews was
under the control of the investigators, and, with
infrequent exceptions, interviews were com-
pleted successfully according to schedule. We

47
Effects of Continued Care

present the number of interviews completed by
experimental groups and by the eleven interview
occasions in Tables 5.1 and 5.2. Each of the 150
patients in the observed group was scheduled for
interviews on 11 separate occasions; intake,
postdischarge, 8 interval interviews by the
regularly assigned interviewer, and a terminal
interview by the newly assigned terminal
interviewer. Of the 1650 (11 x 150) scheduled
interviews for the observed group, 78 per cent
were completed: 17 per cent involved patients

who had died and were, thus, not eligible for
interviews: 4 per cent were partially completed:
and less than 1 per cent were not done. Each of
the 150 patients in the non-observed group was
scheduled for an intake interview and a terminal
interview. Of the 300 (2 x 150) scheduled
interviews for the non-observed group, 85 per
cent were completed: 14 per cent involved
patients who had died and were, thus, not
eligible for interviews: and 1 per cent were not
done or were partially completed.

Table 5.1—-Completeness of Interviews for Observed Subgroups

 

 

 

 

Observed and Observed and
. nursed subgroups (N+0+)! non-nursed subgroups (N-O+)!
Interview
occasion Interviews Interviews Not Interviews Interviews Not
completed not completed eligible completed not completed eligible

Intake 75 0 0 75 0 0
Postdischarge 75 0 0 75 0 0
3 mos. 69 1 5 70 0 5
6 mos. 66 0 9 64 3 8
9 mos. 61 1 13 59 “4 12
12 mos. 57 2 16 59 3 13
15 mos. 54 3 18 56 5 14
18 mos. 51 4 20 51 8 16
21 mos. 48 4 23 51 6 18
24 mos. 44 6 25 42 12 21
Terminal 41 9 25 45 9 21

TOTALS 641 30 154 647 50 128

 

 

 

 

 

 

 

 

 

“Interviews Not Completed” includes interviews partially completed and interviews not done. The column headed “Not Eligible”
includes patients who had previously died and were, thus, not eligible for interviews. The symbols “N+O+> and “N-O+" refer to the

experimental subgroups which had quarterly scheduled interviews as described in Chapter 2.

Table 5.2—Completeness of Interviews for Non-Observed Subgroups

 

 

 

 

 

 

 

 

 

 

 

Non-observed and Non-observed and
Interview nursed subgroups (N+0-)! non-nursed subgroups (N-0-)!
occasion Interviews Interviews Not Interviews Interviews Not
completed not completed eligible completed not completed eligible
Intake 75 0 0 75 0 0
Terminal 50 2 23 54 3 18
Totals 125 2 23 129 3 18

 

 

L“Interviews Not Completed” includes interviews partially completed and interviews not done. The column headed “Not Eligible”
includes patients who had previously died and were, thus, not eligible for interviews. The symbols “N+O-" and “N-O-"" refer to the ex-
perimental subgroups which did not have quarterly scheduled interviews as described in Chapter 2.

48
MEASURES

We collected information for the following
research purposes: (1) to describe the study
sample; (2) to check the comparability of
treatment and control groups; (3) to measure
and describe aspects of the medical and nursing
care; (4) to study the effectiveness of treatment;
and (5) to develop new knowledge about care
and about methods for studying care. Physical,
psychological, and social information about the
patients was collected in terms of the measures
which are summarized in Figure V-1. We
describe the measures in condensed form here
and refer the reader to the sources identified in
Figure V-1 for more detailed definitions and
methods. Chapter 13 of this report includes a
time schedule of evaluations, reproductions of
the forms on which information was recorded,

Interviewing and Measures

and, for certain measures, additional methodo-
logic detail.

As they entered the study, we classified
patients in terms of age, sex, marital status,
education, occupation, Hollingshead’s Two Fac-
tor Index of Social Class, household composi-
tion, types of persons who gave assistance,
recent personal loss, and recent change in
role—as well as in terms of physical, psycholog-
ical, social, and economic function. In each
instance, also, the hospital’s medical staff
determined the principal diagnosis leading to
admission after a complete medical evaluation.
As an index of the presence or absence of
coexisting or associated chronic diseases, every
patient was screened for abnormalities of
electrocardiogram, blood sugar level, blood urea
nitrogen level, cardiac sounds, blood pressure,
weight, vision, use of digitalis, and use of insulin

 

Measures to characterize the population
of chronically ill

Measures to identify changes in function
of the chronically ill

 

 

Age

Sex and race

Primary diagnosis

Coexisting disease (2)

Activities of daily living (Index of ADL) (3)
Walking (2)

Range of movement and strength (4)

House confinement

Marital status

Occupation

Education

Social class (Hollingshead) (5)

Census tract characteristics (6, 7, 8)

Economic resources and productivity (9)

Household composition

Identity of people who assist

Recent personal loss or change in role

Social interaction frequency (adapted Townsend) (10)

Intellectual function (Raven test) (11)
Memory and mental control (adapted Wechsler) (12)
Scale of psychosocial adjustment (Highland View) (13)

 

Death

Activities of daily living (Index of ADL) (3)
Walking (2)

Range of movement and strength (4)

House confinement

Duration of noninstitutional living

Hospitalization

Admission to nursing home

Economic dependence (adapted from Index of ISF) (9)
Social interaction frequency (adapted Townsend) (10)

Intellectual function (Raven test) (11)
Memory and mental control (adapted Wechsler) (12)
Scale of psychosocial adjustment (Highland View) (13)

 

 

INumbers correspond to references at end of chapter.

Figure V-1. Measures in Continued Care Study!

49
Effects of Continued Care

or one of the oral drugs for diabetes. A
systematic review and record was made of the
presence or absence of the following: chest pain
for which a physician advised bed rest for two or
more weeks; a history of a diagnosis of coronary
attack or angina; cancer discovered within five
years by documented record; and inflammatory
or noninflammatory polyarthritis during the
prior two months.

Measures of Physical Function and Injury,
and Measures at Death

We used measures of patient function both to
describe patients as they entered the study and
as criteria of effectiveness of the treatment
program. Measures of activities of daily living
(ADL), walking, house-confinement, and range
of movement and strength had been developed
during 14 years of study in Abington House (2,
3,4,9, 14, 15).

We used the Index of Independence in
Activities of Daily Living, an ordered 7-grade
scale ranging from a high of 4 to a low of G, to
rate functional performance in activities of daily
living (3). Ratings according to this standardized
composite measure were based on detailed
evaluation by both history and observation of
functional dependence in bathing, dressing,
going to toilet, transferring, continence, and
feeding. This index reflects the adequacy of a
spectrum of organized neurological and loco-
motor responses and serves as a graded measure
of such biological and psychosocial function.
Walking was observed and classified according to
the following S-grade scale (2): walking by self,
walking with mechanical assistance, walking
with personal assistance, walking with mechani-
cal and personal assistance, and not walking at
all. We measured house-confinement in terms of
the number of days on which the patient had
been outside of his residence during the 2-week
period which preceded the date of evaluation.
Additional details about the definitions, pro-
cedures, and grading of activities of daily living,
walking, and house-confinement are described in
Chapter 13.

A standardized test of range of movement and
strength was used to define the severity of
neuromuscular disability (4). In this test, we

50

asked the patient to perform a series of six
maneuvers with each lower limb and 10
maneuvers with each upper limb. All maneuvers
were tested both with and without resistance.
Maneuvers were tested using defined positions
for evaluating movement, defined points for
applying resistance, defined avoidance of substi-
tute maneuvers, and standardized definitions of
range of movement in degrees. We classified
function according to this test into four grades
of function for each lower limb and three grades
for each upper limb. Additional details about
the test procedures and grading process are
described in Chapter 13, which includes also the
test form.

Whenever a patient fractured a bone during
the study, we recorded the date and type of
fracture. The occurrence of fracture served as a
measure of injury.

When a study patient died, the following
information was copied directly from the death
certificate; date of death, birthdate, immediate
cause of death, clinical code of cause of death,
underlying cause of death, significant contribu-
ting conditions, presence or absence of autopsy,
place of death (hospital or not), and attendance
by a physician on the date of death. We made no
effort to reconstruct the patient’s functional
status immediately before death because such
information could not be reliably compared
with measurements made at planned intervals.
For deceased patients who had been observed
quarterly, we had complete information through
the last actual interview before death. For those
who had not been observed regularly, interval
functional evaluations were not available.

Some kinds of objective information could be
verified and were obtained for all who died.
These data included date of retirement or
cessation of homemaking; dates of admission to
and discharge from hospitals, nursing homes or
other institutions; whether or not the subject
was cared for at home by a nurse during the
study; and dates of any fractures.

Social and Economic Measures

We recorded social factors in three categories:
(1) socioeconomic status and function, (2) so-
cial interaction, and (3) social deprivation.
Socioeconomic class was measured first by
Hollingshead’s Two Factor Index of Social
Position, which is based on education and
occupation (5). In addition, we used a second
measure which was designated as the Index of
Economic Dependence and which had been
derived from earlier studies on similar groups of
patients (9). Based on detailed evaluations of
indicators of resources and productivity, this
measure consists of four classes in order of
increasing economic dependence, derived from
information about employment status, home
ownership, and receipt of economic assistance
from a charitable agency. Additional details
about the grading process are described in
Chapter 13, which includes also the form used
to record information.

As a measure of social role, we classified
patients according to whether or not they had
been working or acting as homemakers until the
onset of the illness which led to hospitalization.
Dates of cessation were recorded if patients were
no longer employed or homemaking. Household
composition was identified as one of eight
categories: married pair living with children;
married pair living with others; married pair
living alone; solo (i.e., single, separated,
widowed, or divorced) living with children; solo
living with other relatives; solo living with
non-relatives; solo living alone; and not in a
household (usually meaning in a nursing home).
For analysis and as a more clearly graded
measure of household composition, we con-
densed these eight categories into three: married
and living with spouse, solo living with others,
and living alone.

Utilizing certain ideas of Townsend, we
defined a measure of personal interaction based
on the number of kinds of persons with whom
the subject reported having daily contact (10).
The kinds of persons were classified as spouse,
relatives, or friends; and each individual was
identified as having daily contact with 3, 2, 1, or
none of these categories. As other measures of
social activity, we sought the following informa-
tion: the number of people talked to daily,
whether the subject had been employed during
the two weeks before hospitalization, and
whether the subject had engaged in any regular

Interviewing and Measures

organized social activity such as church or club
activities.

We defined social deprivation as one or more
of the following events in the two years before
the subject entered the study: loss of spouse,
retirement, or cessation of homemaking (16).
An additional measure possibly related to these
three, was whether or not the individual
reported that he talked to fewer persons at the
time the study began than he had at age 45.
Institutionalization was recorded in terms of
dates of entry into hospitals and nursing homes,
and the corresponding dates of discharge.

Psychological Measures

Tests of mental and psychosocial function
included the Raven Coloured Progressive Matri-
ces, a standardized memory and mental control
scale adapted from Wechsler, and a Q-sort
method for evaluating psychosocial adjustment
(11, 12,13). The Raven Coloured Progressive
Matrices is a standardized measure of observa-
tion and clear thinking which has been used for
the elderly and which has been found to
correlate reasonably well with the Wechsler
Adult Intelligence Scale (11, 17, 18). This test
consists of three sets of twelve untimed multiple
choice problems, each set being arranged in
order of progressive difficulty. Each problem
consists of a matrix design from which a
segment has been removed plus six possible
inserts from which the subject selects the one
correct, matching insert. Requiring no manual
dexterity and no spoken language, it is suitable
for administration to children, the aged, and the
disabled; and its spectrum of intellectual
difficuly is appropriate for these groups.

We evaluated orientation and mental control
with the aid of an instrument which was adapted
from the Wechsler memory scale (12). This scale
taps the basic mental status of the subject,
testing for orientation for time, place, and
person, as well as for simple mental control. The
orientation sections include questions about the
year, the month, the day, the subject’s physical
location, name, age, and marital status. Evalu-
ation of mental control is based on the subject’s
ability to recite the alphabet, count backwards,
and count by threes.

51
Effects of Continued Care

Interviewers rated the psychosocial adjust-
ment of patients by a Q-sort technique which
had been developed at Highland View Hospi-
tal (13). By this method, interviewers rated 4
dimensions of behavior in each of 3 areas,
namely, motivation, social adjustment, and
personal adjustment. Each descriptive statement
of the rating continuum of each dimension of
behavior was printed on a card. To rate a given
patient, all cards were arranged in order, one end
of the ordered series being those behavioral
characteristics which were most unlike the
patient and the other end being those most like
the patient. Details about Q-sort items, forms,
and scoring are found in Chapter 13.

Measures of Care

Interviewers recorded information about care in
terms of hospitalizations and nursing home
utilization, as well as in terms of physician,
nurse, and other professional services. Dates of
entering and leaving hospitals and nursing homes
were listed and verified by checking against
hospitalization insurance records and, systemati-
cally, with the institutions themselves. Numbers
of contacts with physicians and nurses were
recorded on the basis of questions adapted from
the U. S. National Health Survey (19). Counts
were also made of services by dentists,
optometrists, podiatrists, social workers, and
physical therapists. Nurses of the Visiting Nurse
Association kept records of their services in
relation to exercises, prescribed treatments,
activities of daily living, health promotion,
planning, coordination, and psychological
therapy. Systematic records of these services
were returned to the research offices each
month.

References

1. Kahn, R. L. and Cannell, C. F.: The Dynamics of
Interviewing. Published by John Wiley & Sons, Inc.,
N.Y.,N.Y., pp. 3-130 and 203-232, 1963.

2. Katz, S., Ford, A. B., Heiple, K. G. and Newill, V.
A.: Studies of illness in the aged: Recovery after
fracture of the hip. J. Gerontology 19: 285-293,
1964.

52

10.

11.

12.

13.

14.

16.

. The

. Katz, S., Ford, A. B., Moskowitz, R. W., Jackson, B.

A., Jaffe, M. A.: Studies of illness in the aged: The
index of ADL, a standardized measure of biological
and psychosocial function. J A.M. A. 185: 914-919,
1963.

. Katz, S., Ford, A. B., Chinn, A. B. and Newill, V.

A.: Prognosis after strokes. Part II. Long-term
course of 159 patients. Med. 45: 236-246, 1966.

. Hollingshead, A. B.: Two Factor Index of Social

Position. Published by Hollingshead, New Haven,
Conn., pp. 1-11, 1957.

. U. S. Department of Commerce, Bureau of the

Census: United States Census of Housing, 1960,
Cleveland, Ohio Area (Final Report HC(2) 45).
Published by U. S. Government Printing Office,
Washington, D.C., pp. 1-30, 1962.

. Finley, J. E.: Greater Cleveland and Its Health.

Published by the Cleveland Welfare Federation,
Cleveland, Ohio, 1963.

. Cleveland Real Property Inventory: Census tract

street index for Cuyahoga County and Western Lake
County. 7th Ed. Published by Real Property
Inventory of Metropolitan Cleveland, Cleveland,
Ohio, pp. 1-73, 1962.

Staff of The Benjamin Rose Hospital:
Multidisciplinary studies of illness in aged persons.
V. A new classification of socio-economic function-
ing of the aged. J. Chron. Dis. 13: 453464, 1961.

Townsend, P.. The Family Life of Old People.
Published by Routledge & Kegan Paul, Ltd.,
London, England, pp. 166-182, 1957.

Raven, J. C.: Coloured Progressive Matrices (Sets A,
Ag , B of revised order 1956). Published by H. K.
Lewis & Co., Ltd., London, England, 1962.

Wechsler, D.: A standardized memory scale for
clinical use. J. Psychol. 19: 87-95, 1945.

Shontz, F. C. and Fink, S. L.: A method for
evaluating psychosocial adjustment of the chroni-
cally ill. Am. J. Phys. Med. 40: 63-69, 1961.

Katz, S., Downs, T. D., Cash, H. R. and Grotz, R.
C.: Progress in Development of the Index of ADL.
Gerontologist 10: 20-30, 1970..

. Steinberg, F. U. and Frost, T.M.: Rehabilitation of

Geriatric Patients in a General Hospital: A follow-up
study of 43 cases. Geriatrics 18: 158-164, 1963.

Shanas, E., Townsend, P., Wedderburn, D., Friis, H.,
Milhgj, P. and Stehouwer, J.: Old People in Three
17.

18.

Industrial Societies, First Edition. Published by
Atherton Press, N.Y, N.Y, pp. 258-287, 1968.

Martin, A. S. and Wiechers, J. E.: Raven’s Coloured
Progressive Matrices and the Wechsler Intelligence
Scale for Children. J. Consult. Psychol. 18:
143-144, 1954.

Kidd, C. B.: Criteria for admission of the elderly to

19.

Interviewing and Measures

geriatric and psychiatric units. J. Ment. Sci. 108:
68-74, 1962.

U.S. Dept. of Health, Education, and Welfare: U.S.
National Health Survey Health Statistics, Public
Health Service Publication No. 584-B 10. Published
by U. S. Government Printing Office, Washington,
D.C., pp 66-67, 1959.

53
CHAPTER VI.

CLASSIFICATION OF DISEASE, DISABILITY

AND MENTAL STATUS AT INTAKE

The subjects of this study had moderate to
severe physical disability. A few showed mental
confusion, and all had been admitted to the
hospital at one stage or another of various
chronic diseases (Chapter 3). Each of these
characteristics, alone or in combination with
others, was likely to exert its particular
influence on the course to be expected, and on
the person’s capacity to respond to the nursing
program. The number of characteristics and
combinations was large; thus, the population
was quite heterogeneous in terms of disease,
disability, and mental status. Since hetero-
geneity is associated with increased variability
which could obscure many effects of the
treatment program, we found it necessary to
establish subgroups of individuals who were
comparable at the time they entered the study.
This chapter describes the systems of classifica-
tion which we used to establish comparable
subgroups for disease, disability, and mental
status; and the next chapter (Chapter 7) deals
with the treatment program’s effects in terms of
these classification systems. Classes were defined
before results were analyzed in order to avoid
the biases of retrospective classification.

Our approach to classification had been
developed during a period of about 13 years
through epidemiological and experimental
studies of chronically ill patients. The following
discussion is based on the experiences in these
past studies and serves as important background
information with respect to the approach to
classification which we took in the present
study.

Information which can be used to classify
disease, disability, and mental status can be
obtained from the customary procedures of

54

clinical medicine or from systematic application
of predefined criteria by means of questions,
observations, or tests. Classes can, thereby, be
defined in such terms as clinical diagnosis or as
signs or indicators of disease or disability. Use of
terms of clinical diagnosis as criteria for
classifying patients is illustrated by the Interna-
tional Classification of Diseases, which is based
on a mixture of anatomic, clinical, etiologic,
pathologic, and physiologic criteria (1). Use of
indicators of disease and disability to classify
population groups is illustrated in reports
concerning the prevalence of disease and
disability by the National Health Survey and
Commission on Chronic Illness (2, 3,4). Each
type of information has advantages and
disadvantages, and neither is adequate by itself
to define groups of persons with comparable
prognoses or to establish groups which are
ordered according to severity of illness.

Diagnosis is a common basis for classification
and is the usual starting point in defining
patients’ needs for care and in organizing
services. Problems encountered with existing
systems of diagnostic classification relate to
difficulties in establishing specific diagnoses,
difficulties in defining the stage and severity of
chronic illness, and the great variability of
chronic disease characteristics. With respect to
diagnosis, we lack the means to identify with
certainty the existence and stage of many
chronic diseases as, for example, in certain early
stages of hypertensive cardiovascular disease,
arteriosclerotic heart disease, cancer, or mental
disorder (5, 6, 7, 8, 9). In the absence of
knowledge about cause, diagnostic procedures
are not completely definitive; and we miss
numerous instances of early and mild illness.
Classification of patients according to clinical
diagnosis does not, therefore, include all those
who have a specific disease. Attempts to search
for the presence or absence of all diagnoses are
impractical, and certain diagnostic procedures
involve risks to patients. Because of variations in
skill, information, and criteria, the use of
diagnostic categories alone to describe large
populations is likely to result in long lists with
many small subgroups and much overlapping.
Associated with incomplete knowledge about
causes of disease and pathogenesis is a dearth of
quantitative information about the natural
progression and regression of chronic dis-
ease (10). As a result, knowledge about mile-
stones of disease is poorly developed; and
existing systems of diagnostic classification are
of little value to describe changes in the stages
and severity of illness. Finally, current systems
of classification do not adequately take into
account the marked variation in patterns of
chronic disease in terms of the kinds, numbers,
severity, and complications of disease. Two
patients, both with the diagnosis of stroke, for
example, may differ greatly in the amount of
measurable disability they present. On the other
hand, two patients with the same degree of
disability may have quite different prognoses if
the disability is the result of different disease
conditions. In a study such as the present one, a
system of classification which does not recog-
nize variation in disease patterns is likely to
obscure study results.

In contrast to diagnostic classification, the use
of signs or indicators permits systematic
characterization of population groups in stand-
ard terms and saves time and cost. (Nonsys-
tematic diagnostic review, no matter how
thorough, does not result in comparable classes).
Despite such advantages, however, the use of
indicators for classification has distinct limita-
tions. Indicators are selected on the basis of
their validity as indicators of disease. Since our
current knowledge about causes and dynamics
of chronic disease is generally incomplete,
indicators do not establish with certainty
whether the respective conditions are present or
absent. For example, the sensitivity and
specificity of indicators are unknown with

Disease, Disability and Mental Status at Intake

respect to most chronic diseases. As a result, the
occurrence of false positives and false negatives
could obscure interpretations about findings in a
study such as ours. In this type of study, also,
only indicators of fairly prevalent conditions can
be used.

Classification of patients according to disabil-
ity has certain advantages over classification
according to clinical diagnoses or indicators of
disease. As a criterion which can be measured
objectively, disability is an index of the
existence, stage, and effects of chronic disease at
a time when information about etiology and
pathogenesis is not complete enough to permit
measurement in more basic terms. Degree of
disability is, thus, one index of severity of illness
and a sensitive indicator of changes in the course
of illness. An important advantage offered by
measures of disability and function is that they
provide a conceptual basis for homogeneous
groupings in diagnostically heterogeneous popu-
lations. Examples of the use of disability
classifications as measures of morbidity are
found in reports of the Commission on Chronic
Illness and National Health Survey, which
describe the prevalence of various types of
disability in large populations (2, 3, 11). In the
present study, we have made use of a measure of
disability, the Index of Independence in
Activities of Daily Living, which we developed
in extensive studies of similar patients (12, 13).
Its reliability and validity have been studied, and
it has been shown to differentiate groups of
patients with different prognoses and to be
related to patterns of physical and behavioral
development (12, 13, 14, 15).

In view of the previously discussed problems
and concepts, we established three systems of
classification from combinations of three types
of information in order to study the effects of
the treatment program in comparable subgroups
of individuals. The three types of information
were: principal clinical diagnosis, indicators of
chronic conditions, and degree of disability.

Each type of information served as the basis for
a system of classification, and the three systems

will be described below. The classifications were
based on previous experiences in studies of the
natural history of stroke, fracture of the hip,

55
Effects of Continued Care

and Parkinson’s disease (14, 15, 16); and they
had also been used successfully in a controlled
study of comprehensive care for outpatients
with rheumatoid arthritis (17). An important
criterion in selecting classes was to have them
contain sufficient numbers of subjects to permit
meaningful analysis.

Classes Based on Principal Clinical Diagnosis

For the first system of classification, we grouped
patients according to principal diagnosis, defined
as the diagnosis which led to the hospital
admission (Chapter 3). We divided each diag-
nostic class into two subcategories, one contain-
ing those who had the specific diagnosis plus one
or more indicators of coexisting major chronic
conditions (cardiovascular-renal disease, cancer,
severe disability, or dementia), and the other
containing persons with the same diagnosis but
without these indicators of chronic conditions.
Since the indicators were measured in similar
terms for all subjects, the subdivided classes
were more homogeneous than groups of patients
classified according to principal diagnosis alone.

Stroke and fracture of the femur were the
two diagnostic categories which contained the
largest numbers of individuals, while other single
disease classes were not large enough for analy-
sis. By including closely related diagnoses, we
defined enlarged diagnostic classes and subdi-
vided them according to the presence or absence
of indicators of coexisting major chronic condi-
tions. The combined categories included brain
disease, muscular and skeletal conditions, and
cardiovascular-renal disease. We established a
total of fourteen classes as defined in Table 6.1,
which also defines the indicators of coexisting
major chronic conditions (Intake Classes #28
through #42 in Chapter 14). Excluded from
Table 6.1 are the classes which were not large
enough for analysis, either as single diseases or as
combined categories.

Classes According to
Disease-Disability Profiles

The second type of classification grouped pa-
tients according to profiles of disease and

56

Table 6.1 Classes Based on Principal Diagnosis
and Coexisting Major Chronic Conditions

 

 

oy . . Coexisting No. of
P Pp
rincipal Diagnosis condition! persons
Stroke Present 60
Stroke Absent 15
Stroke or chronic Present 76
brain syndrome
Stroke or chronic Absent 18
brain syndrome
Any brain condition? Present 90
Any brain condition Absent 21
Fraction of femur Present 34
Fraction of femur Absent 23
Fracture of lower limb3 Present 4]
Fracture of lower limb Absent 28
Any muscular or skeletal Present 57
condition4
Any muscular or skeletal Absent 53
condition
Cardiovascular-renal Present 77
disease’
Cardiovascular-renal Absent 24
disease

 

 

 

 

 

coexisting chronic conditions are one or more of: 1) cardio-
vascular-renal disease (abnormal Q-wave on EKG, or blood urea
nitrogen of 30 mg. or more per 100 ml.), 2) cancer occurring
within § years, 3) severe general disability (Index of ADL grade
of F or G), or 4) dementia (Raven test score of 12 or less, or
orientation test score of 13 or less).

Stroke, chronic brain syndrome, paralysis agitans, multiple
sclerosis, subarachnoid hemorrhage, subdural hematoma, cere-
bral contusion, idiopathic cerebellar atrophy, idiopathic pro-
gressive paresis, basilar artery insufficiency, or apraxia.
Fractures of femur, femoral condyle, tibia, fibula, ankle, or
metatarsals. A patient with partial ankylosis of the knee after
fracture and with posttraumatic subluxations of metatarsals
were also included.

4Fractures of the lower limb (defined above), osteoarthritis,

rheumatoid arthritis, idiopathic myopathy, gout, chondrocal-
cinosis, ligamentous strain of hip, fractured clavicle, fractured
humerus, Colles’ fracture, fractured wrist, osteoporotic com-
pression fractures of vertebrae, fractured pelvis, aseptic necro-
sis of head of femur, and atrophy after poliomyelitis with post-
fracture nonunion and amputation.

Arteriosclerotic heart disease (including myocardial infarction),
stroke, generalized arteriosclerosis, hypertensive cardiovascular
disease, rheumatic heart disease, basilar artery insufficiency,
anterior spinal artery syndrome, and amputations for occulu-
sive vascular disease.
disability, based on an idea developed by mem-
bers of the Commission of Chronic Illness (11).
By this system, persons in the Commission’s
study of chronic illness were divided into three
classes. The first class included those with
maximum disability regardless of diagnosis. The
second and third classes had less disability, the
second class having some disease and the third
having no or minor disease. The three disease-
disability classes were, thus, ordered according
to severity and reflected the burden of illness in
those who were evaluated.

In the present study, we divided all 300
patients into three classes according to the above
concept. Those with maximum disability (having
an Index of ADL grade of FF or G) were included
in the first class regardless of diagnosis (Intake
Class #2 in Chapter 14). Patients in the second
and third classes had less disability. Those in the
second class had at least one indicator of
coexisting major chronic conditions (Intake
Class #3 in Chapter 14), while those in the third
class had no such indicators (Intake Class #4 in
Chapter 14). The use of indicators, instead of
diagnoses as used in the Commission studies,
took advantage of the better reliability of
systematically tested disease indicators. The
coexisting chronic conditions for the second
class were cancer (documented occurrence
within 5 years), coronary disease (abnormal
Q-waves on EKG), and renal disease (blood urea
nitrogen level of 30 mg. or more per 100 ml.).

We evaluated the effect of varying definitions
for the second class by adding one or more of
the following indicators: hypertension, mainte-
nance therapy with digitalis, and proteinuria.
Based on such modified definitions, we estab-
lished four other sets of classes, the respective
four sets being Intake Classes #2, #5, #6, #7, #8;
#2, #9, #10; and #2, #11, #12. Patients were
distributed approximately in thirds between the
classes of each set, no class having fewer than 74
or more than 133 patients.

Classes Based on Indicators of
Major Chronic Conditions

‘The final type of classification grouped patients
according to patterns of four major chronic
conditions, namely, cardiovascular-renal disease,

Disease, Disability and Mental Status at Intake

cancer, severe disability, and dementia (Intake
Classes #13 through #27 in Chapter 14). The
first two conditions (cardiovascular-renal disease
and cancer) identify the leading causes of death
in the United States today, while the latter
conditions (disability and dementia) reflect
physical and mental morbidity which is most
serious for the particular patients in this study.
By the term “dementia”, we mean ‘any con-
dition of deteriorated mentality” which is
roughly synonymous with the common term
“senility” or the clinical diagnosis of chronic
brain syndrome (18).

By systematic evaluation, we established
whether the previously defined indicators of
these conditions were present or absent, and we
produced comparable information for all pa-
tients (Tables 3.2, 3.3, and 3.6). Cutoff points
with respect to these indicators were defined in
relation to their relevance to the people and
purposes of this particular study. To the degree
that knowledge is limited, we recognize that
judgments differ about cutoff levels and defini-
tions. The conceptual bases and definitions in
this study are, however, presented clearly and
will permit the reader to visualize the basis for
interpretation of study results and the people to
whom such results apply.

References

1. US. Dept. of Health, Education, and Welfare:
International Classification of Diseases, Adapted.
Public Health Service, publication No. 719, Vol. 1
& 2. Published by the U.S. Government Printing
Office, Washington, D.C., 1962.

2. U.S. National Center for Health Statistics: Chronic
Conditions Causing Limitation of Activities, United
States, July 1959-June 1961. U. S. Public Health
Service, publication No. 584-B36, Series B-No. 36,
U. S. Dept. of Health, Education, and Welfare,
Washington, D. C., 1962.

3. U. S. National Center for Health Statistics:
Disability Days Due to Injury, United States, July
1959-June 1961. U. S. Public Health Service,
publication No. 584-B40, Series B-No. 40, U. S.
Dept. of Health, Education, and Welfare, Washing-
ton, D.C., 1963.

4. Trussel, R. E. and Elison, J.: Chronic Illness in the
United States, Vol. Ill, Chronic Illness in a Rural

57
Effects of Continued Care

Area. Published by Harvard University Press, Cam-
bridge, Mass., pp. 225-308, 1959.

. Clark, E. G., and Morsell, J. A.: An Epidemiologic

Approach to the Study of High Blood Pressure. Am.
J. Pub. Health 42: 542-548, May 1952.

. Commission on Chronic Illness. Chronic Illness in

the United States, Vol. Il, Care of the Long-term
patient. Published for the Commonwealth Fund,
Harvard University Press, Cambridge, Mass., pp.
332-341, 1956.

. Drake, R. M., Buechley, R. W., Breslow, I. L.: An

Epidemiological Investigation of Coronary Heart
Disease in the California Health Survey Population.
Am. J. Pub. Health 47: 43-57, April 1957 (Supple-
ment, Part II).

. Gordon, J. E., O'Rourke, E., Richardson, F. L. W.

Jr. and Lindemann, E.: The Biological and Social
Sciences in an Epidemiology of Mental Disorder.
Am. J. Med. Sci. 223: 316-343, March 1952.

. The National Health Education Committee, Inc.:

Facts on the Major Killing and Crippling Diseases in
the United States Today. Published by the National
Health Education Committee, Inc., N.Y. N.Y,
1959.

10. Dyar, R.: Problems in the Measurement of the

58

Progression and Regression of Chronic Disease.
Milbank Mem. Fund Quart. 31: 239-241, July 1953.

11.

12.

13.

14.

15.

16.

17.

18.

Commission on Chronic Illness: Chronic Illness in
the United States, Vol. IV, Chronic Illness in a
Large City. Published by Harvard University Press,
Cambridge, Mass., pp. 375-377, 1957.

Katz, S., Ford, A. B., Moskowitz, R. W., Jackson, B.
A. and Jaffe, M. W.: Studies of Illness in the
Aged—The Index of ADL: A Standardized Measure
of Biological and Psychosocial Function. J. 4. M. A.
185: 914-919, 1963.

Katz, S., Downs, T. D., Cash, H. R., and Grotz, R.
C.: Progress in Development of the Index of ADL.
The Gerontologist 10: 20-30, 1970.

Katz, S., Ford, A. B., Chinn, A. B. and Newill, V.
A.: Prognosis after Strokes, Part II, Long-Term
Course of 159 Patients. Med. 45: 236-246, 1966.

Katz, S., Heiple, K. G., Downs, T. D., Ford, A. B.
and Scott, C. P.: Long-Term course of 147 Patients
with Fracture of the Hip. Surgery, Gynecology and
Obstetrics 124: 1219-1230, 1967.

Steinberg, F. U. and Frost, T. M.: Rehabilitation of
Geriatric Patients in a General Hospital: a follow-up
study of 43 cases. Geriatrics 18: 158-164, 1963.
Katz, S., Vignos, P. J., Jr., Moskowitz, R. W.,
Thompson, H. M. and Svec, K. H.: Comprehensive
Outpatient Care in Rheumatoid Arthritis: a Con-
trolled Study. J. 4. M. A. 206: 1249-1254, 1968.

Webster’s Collegiate Dictionary, 5th Ed., 1937.
CHAPTER VII.

EFFECTS OF TREATMENT PROGRAM: ALL PATIENTS

AND SUBGROUPS BY DISEASE, DISABILITY AND MENTAL STATUS

In the preceding chapter, we described a system
of disease and disability classification which had
been developed to define homogeneous sub-
groups of patients. The concepts and experi-
ences on which the syntheses were based were
also described in detail. The present chapter
reports significant associations between the
treatment program and the outcome for sub-
groups of patients which were homogeneous for
principal clinical diagnosis: (with and without
coexisting chronic conditions), disease-disability
profiles, and indicators of major chronic condi-
tions (cardiovascular-renal disease, cancer, severe
disability, and dementia). The first section of
the chapter deals with effects of the nursing
program in terms of function (physical, mental,
and social), injury, and mortality. The second
section discusses outcome in terms of use of
services.

Areas of Measurement

Our primary interest in the study was in the
effectiveness of the treatment program in main-
taining or improving physical, psychological, and
social well-being of the patients (measured in
terms of function). Certain kinds of interpre-
tations about patient well-being are, of course,
more acceptable than others, due to limitations
of available measures. For some functions, for
example, such as walking and activities of daily
living, we have precisely defined reproduci-
ble measures, while for other functions, such as
social interaction and social adjustment, defini-
tions and reproducibility of the measures are
limited. Another limitation is that currently
available measures do not cover all the areas of
function included in the concept of compre-
hensive human well-being.

Our second interest in the study was in the
influence of the treatment program on patients’

use of medical and other services. To the extent
that use of services indicates that patients are
receiving the benefit of the skills and experi-
ences which are part of the professional’s back-
ground, such use is an indicator of quality care.
One must bear in mind, however, that this
consideration is theoretic; and the actual effect
on patient well-being is not necessarily proved
by demonstrating increased use of services alone.
At the same time, the possibility exists that, for
patients who are chronically ill, use of services
reflects the public health nurse’s coordinator
function and is definitely related to the patients’
welfare.

In presenting the results of the study, we shall
represent both of the areas of interest men-
tioned above. We shall describe the first area
under the heading Function, Injury, and Mortal-
ity, and the second under the heading Use of
Services. Function was evaluated in terms of
activities of daily living, walking, range of
movement and strength, orientation and mental
control, observation and clear thinking, degree
of house-confinement, socioeconomic function,
and social interaction, while the occurrence of
injury was evaluated in terms of the occurrence
of fractures (Outcome Classes #3 through #17
in Chapter 14). We evaluated use of services by
patients in the study in terms of hospitaliza-
tions, admissions to nursing homes, nurse visits,
physician visits, services by other professionals
and nonprofessionals, and service circumstances
at the time of death, including attendance by a
physician on the day of death, death in hospital,
and autopsy (Outcome Classes #18 through #28
in Chapter 14).

For the analyses in both this chapter and
Chapter 9, definitions of change in physical,
mental, and social function excluded death as an
item of change. In relation to physical and

59°
Effects of Continued Care

mental function, a ‘beneficial’ treatment effect
was defined as improvement or maintenance
(not improvement alone), since the study’s main
hypothesis was that function would be improved
or maintained in the presence of the experi-
mental care variable. We did not, however, use
the terms “benefit” or “improvement” to de-
scribe social effects, since changes in social
functions as, for example, participation in social
activities, cannot be clearly interpreted as
changes for the better or worse in this disabled
and aging population. Such changes in social
function are less clearly measures of well-being
than are changes in physical and mental func-
tion. The same is true for changes in house-
confinement which may be influenced as much
by social and psychological factors as by physi-
cal disability. We, therefore, described changes
in social function and house-confinement as
increases or decreases; and interpretation of the
meaning of such increases or decreases are
explored in Chapter 11, where we discuss
implication of the study results. Lastly, in both
this chapter and in Chapter 9, such phrases as
“more often associated”, ‘“‘occurred more fre-
quently”, “more frequently maintained or im-
proved”, “fewer deteriorations”, ‘beneficial
nursing effect”, “more services” and “fewer
admissions” are all based on demonstrated sig-
nificance by statistical test and on the principles
of statistical inference which are described in
Chapter 12.

Outcome in Terms of Function,
Injury, and Mortality

Differences in function, injury, and mortality were
not demonstrated between the study groups which
included all patients, namely, between the hetero-
geneous nursed and non-nursed groups.

Among certain subgroups which were homogene-
ous for disease and disability, statistically signifi-
cant and consistent differences were demonstrated
between nursed and non-nursed groups. The most
frequent outcomes in the presence of the nursing
program were: (a) beneficial nursing effects in
physical and/or mental function among patients
who had a principal diagnosis of a muscular or
skeletal condition without a coexisting major
chronic condition, those who had a principal
diagnosis of arthritis, those who were least severely
ill, and those with less than maximum disability;

60

and (b) decreased social interaction for a few
subgroups which had, in common, poor mental
function and cardiovascular-renal abnormalities.

By statistical test, we did not demonstrate
differences in function, injury, and mortality
between the study groups which included all
patients, namely, between the heterogeneous
nursed and non-nursed groups. This was true for
analyses of differences in activities of daily
living, walking, range of movement and strength,
orientation and mental control, observation and
clear thinking, socioeconomic function, social
interaction, degree of house-confinement, occur-
rence of fractures during the study, and mortal-
ity.

Among subgroups which were homogeneous
for disease, disability, and mental status, we
demonstrated statistically significant differences
which revealed associations between the treat-
ment program and the study’s outcome. These
associations in homogeneous subgroups will be
described next, and we refer the reader to
Chapter 6 for more detailed definitions of the
classes of disease, disability, and mental func-
tion.

Considered, first, are outcomes for patients
classified according to principal diagnosis as they
entered the study. For patients who had a
principal diagnosis of a muscular or skeletal
condition without a coexisting major chronic
condition (indicators of cardiovascular-renal dis-
ease, cancer, severe disability, or dementia), the
presence of the nursing program was associated
with a beneficial outcome. The types of bene-
ficial outcome included fewer deteriorations by
measures of walking, range of motion, and

orientation and mental control (Table 7.1).
We did not demonstrate differences in the

previously listed outcome measures of function,
injury, and mortality between nursed and non-
nursed subgroups of patients who had a princi-
pal diagnosis of fractured femur or another
fracture of the lower limb, whether such groups
included those with (41 patients) or without (28
patients) a coexisting major chronic condition.
It should be noted, however, that the experi-
mental subgroups in these instances were small.
Patients who had a fracture of the femur or
another lower limb fracture comprised almost
Effects of Treatment Program

Table 7.1—Nursing Effect on Deterioration in Function Two Years After
Study Intake in Patients Who Had Principal Diagnoses of Muscular or
Skeletal Conditions

 

 

 

Patients who had muscular or
skeletal conditions!
Referred to Not referred Totals
Visiting to Visiting
Nurse (N+) Nurse (N-)
Walking
Deterioration? 0 7 7
No deterioration 16 20 36
Range of Movement
Deterioration? 1 11 12
No deterioration 15 16 31
Orientation and Mental Control
Deterioration? 1 10 11
No deterioration 15 14 29

 

 

 

 

 

 

Ipatients include those who had principal diagnoses of muscular or skeletal conditions without coexisting major chronic conditions as
defined in Table 6.4 of Chapter 6.

2Deterioration in walking is Outcome Class #4 defined in Chapter 14. Deterioration in range of movement is Outcome Class #9.
Deterioration in orientation and mental control is Outcome Class #11. Differences in totals reflect variations in the numbers of patients

for whom identification of deterioration was possible.

two-thirds of those who had a principal diag-
nosis of a muscular or skeletal condition (69 of
110 patients). About three-fourths of the re-
maining patients who had muscular and skeletal
conditions had a principal diagnosis of arthritis;
and Table 7.2 presents evidence of the beneficial
effects of the nursing program on the outcome
in patients with arthritis in terms of fewer
deteriorations in range of movement.

We did not demonstrate differences in any of
the 16 outcome measures of function, injury,
and mortality between nursed and non-nursed
subgroups of patients who had a principal
diagnosis of stroke. This was true for those who
had stroke with or without a coexisting major
chronic condition. Similarly, no difference was
observed between nursed and non-nursed sub-
groups of patients classified according to com-
bined categories of central nervous system diag-
noses, namely, stroke or chronic brain
syndrome, any brain condition, and cardio-
vascular-renal disease (including vascular disease
of brain or spinal cord).

We, next, considered outcome in patients
classified according to disease-disability profiles
which were ordered according to severity. Ex-
perimental subgroups with maximum disability
(assisted in bathing, dressing, toileting, and
transferring, in addition to being incontinent
and/or receiving assistance in feeding) did not
differ in outcome by any of the measures of
function, injury, and mortality. Experimental
subgroups with less than the preceding degree of
maximum disability did show differences in
outcome, and such differences were limited to
outcome in physical function (Figure VII-1).
Among patients with less than maximum disabil-
ity and with coexisting major chronic condi-
tions, the subgroup assigned to the nursing
program had fewer deteriorations by the range
of movement test. In this class of patients, 1
deteriorated among 38 assigned to the nursing
program, and 7 deteriorated among 41 not so
assigned. This beneficial nursing effect was
evident for those having one or more of the
following abnormalities: cancer present in the

61
Effects of Continued Care

Table 7.2—Nursing Effect on Deterioration in Range of Movement
Two Years After Study Intake in Patients Who Had Arthritis

 

Patients Who Had Arthritis!

 

 

 

 

Referred to Not referred
Visiting to Visiting Totals
Nurse (N+) Nurse (N-)
Range of Movement
Deterioration? 11 13
No deterioration 15

 

 

 

patients include those who had principal diagnoses of rheumatoid arthritis, osteoarthritis, gout, or pseudogout.
2Deterioration in range of movement is Outcome Class #9 defined in Chapter 14.

past five years, abnormal Q-waves on the electro-
cardiogram, azotemia, or taking a cardiac glyco-
side regularly. When we modified the classifi-
cation of major abnormalities by including
various combinations of other cardiovascular-
renal criteria (hypertension and proteinuria), we
did not demonstrate differences in outcome
between the experimental subgroups.

Experimental subgroups with the least severe
of the three disease-disability patterns, namely,
those without maximum disability and without
coexisting major chronic conditions, also dif-
fered in outcome. Walking and range of move-
ment were more frequently maintained or im-
proved in the nursed group than in the
non-nursed group. The respective ratios in the
nursed and non-nursed groups were 35 of 38
patients and 31 of 41 for walking, and the
respective ratios were 34 of 38 and 31 of 43
for range of movement.

Outcome was also studied in patients classi-
fied according to disability status in terms of
activities of daily living and in the large number
222 patients) with indicators of major cardio-
vascular-renal disease. In these studies, we also
evaluated outcome with respect to adequate or
poor mental‘ function. Patients in the less de-
pendent classes of activities of daily living
(grades A, B, C, or D by the Index of ADL)
more often improved or maintained both range
of movement and orientation and mental con-
trol in the presence of the nursing program than
in its absence (Table 7.3). When patients in these
less dependent classes of activities of daily living

62

were subdivided according to the presence of
adequate or poor mental function, the same
beneficial nursing effect was demonstrated for
those with good mental function, while no such
effect was demonstrated for those with poor
mental function. After demonstrating beneficial
nursing effects among those who had lesser
degrees of disability in activities of daily living,
we evaluated effects among patients classified in
other disability terms. The least disabled pa-
tients consistently benefited the most. Confirm-
atory evidence included fewer deteriorations in
range of movement among those who were less
limited in range of movement at intake and
fewer deteriorations in orientation and mental
control among those who were able to walk
without assistance at intake. The outcome stud-
ies in disability and cardiovascular-renal sub-
groups did not, otherwise, add to or change
interpretations made in the outcome studies
based on other classification systems of disease
and disability described in this chapter.

Finally, for patients in certain disease classes,
a type of social outcome which was more
frequent in the nursed group than in the
non-nursed groups was the outcome which
reflected decreased social interaction and which
was defined as “fewer social activities at the
terminal point than at intake”. Included were 4
classes, 3 of which had, in common, poor mental
function and 2 of which had evidence of
cardiovascular-renal disease. The remaining dis-
ease and disability classes were not associated
with differences in function, injury, or mortality

 
£9

 

Severity of illness grades?

Nursing effect on individual functions

1

 

Range of movement

 

 

 

 

 

 

 

 

Walking Activities of daily living
Upper limb Lower limb Any limb
Most severe 0? 03 03 0? 03
Moderately severe 0? 03 03 03 Beneficial effect
Least severe Beneficial effect 03 Beneficial effect 0? Beneficial effect

 

 

Individual functions are defined in Chapter 5, and the measures of effects are the changes in function expressed as Outcome Classes #3, #4, #7, #8, and #9 in Chapter 14. The nurs-
ing effect was defined as “beneficial” when function was maintained or improved more frequently in the presence of the nursing program than in its absence (by statistical test).
Severity of illness grades are defined in Chapter 6 according to disease-disability profiles and are specified as Intake Classes #2 through #12 in Chapter 14. Patients in Intake Class
#2 were the most severely ill. Those with illness of moderate severity were in any of Intake Classes #3, #5, #7, #9, or #11; and the least severely ill were in any of the Intake Classes

#4, #6, #8, #10, or #12.

30 = no beneficial or adverse effect demonstrated.

Figure VII—1. Treatment Effect on Physical Function According to Severity of Illness

wero JuaweaI] Jo $3093
Effects of Continued Care

Table 7.3—Nursing Effect on Deterioration in Function Two Years After
Study Intake in Patients Graded A, B, C, or D by the Index of ADL

 

 

 

Patients Graded
A,B,C, or D by Index of ADL!
Referred to Not referred Totals
Visiting to Visiting
Nurse (N+) Nurse (N-)
Range of Movement
Deterioration? 1 10 11
No deterioration 26 25 51
Orientation and Mental Control
Deterioration? 6 16 22
No deterioration 20 16 36

 

 

 

 

 

 

lgee Chapter § for detailed definitions of the Index of ADL. By the Index, patients graded as 4, B, C, or D are more independent in
activities of daily living than those graded as E, F, or G.
Deterioration in range of movement is Outcome Class #9 defined in Chapter 14. Deterioration in orientation and mental control is
Outcome Class #11. The difference in totals reflects the difference in numbers of patients for whom identification of deterioration was

possible.

in terms of these or any of the other measures of
physical, psychological, or social outcome be-
tween the nursed and non-nursed subgroups.

Outcome in Terms of Use of Services

Aside from the expected difference of more
visiting nurse visits in the experimental group
which was designed to receive such visits, more
patients in the nursed group were admitted to
hospitals during the study than in the non-nursed
group; and autopsies were performed more fre-
quently on those in the nursed group who died.

Among subgroups which were more homogeneous
for disease and disability at the time of entry into
the study, the most frequent outcomes in the
presence of the nursing program were: (a) more
patients hospitalized among those who had a
principal diagnosis of fracture of the femur plus a
coexisting major chronic condition, those who
were most disabled, and those with less than
maximum disability who were also free of major
chronic conditions; (b) more contacts with physi-
cians among those who had a principal diagnosis of
stroke plus a coexisting major chronic condition,
those who had a principal diagnosis of cardiovascu-
lar-renal disease plus a coexisting major chronic
condition, and those who were most disabled;
(c) more non-nurse and non-physician services
among those who were most disabled and among
those with less than maximum disability who were
also free of major chronic conditions; (d) fewer

64

admissions to nursing homes among patients who

were not maximally disabled in activities of daily

living at the time of entry into the study; and

(e) more hospital services at the time of death

among those who had a principal diagnosis of

cardiovascular-renal disease plus a coexisting major
chronic condition and those who were most
disabled.

Concerning the use of services by patients in
the study, we demonstrated differences between
the study groups which included all patients.
Aside from the expected difference of more
visiting nurse visits in the experimental group
which was designed to receive such service, more
patients in the nursed group (95 of 141 patients)
were admitted to the hospital during the study
than in the non-nursed group (76 of 137
patients). The frequency of hospitalizations is
described in Table 7.4. Autopsies were, also,
more frequently performed on deceased patients
in the nursing program than on those not in the
nursing program.

We devote the rest of this section to outcome
in homogeneous subgroups of the nursed and
non-nursed groups. We shall, first, discuss out-
come in subgroups which were homogeneous
with respect to principal diagnosis of patients.
More patients in the nursed group who had a
fracture of the femur plus a coexisting major
Table 7.4—Nursing Effect on Frequency of
Hospitalizations During the Study

 

 

 

 

Experimental Groups
Number of
Hospitalizations Referred to Not referred
Visiting to Visiting
Nurse (N+) (Nurse (N-)
(number of persons)
0 55 74
1 51 37
2 28 22
3 or more 16 17
TOTALS 150 150

 

 

 

 

 

chronic condition were hospitalized than were
similar patients in the non-nursed group (12 of
16 in the nursed group as compared to 4 of 15
in the non-nursed group). As related and consist-
ent findings, more of those with principal
diagnoses of any lower limb fracture plus major
comorbidity and more with principal diagnoses
of muscular or skeletal condition plus major
comorbidity were hospitalized in the presence of
the nursing program (14 of 21 with any lower
limb fracture and 19 of 28 with a muscular or
skeletal condition) than in its absence (5 of 17
with any lower limb fracture and 10 of 25 with
a muscular or skeletal condition). The reader
should note that, of the 53 with principal
diagnoses of muscular or skeletal condition plus
major comorbidity, 31 had a principal diagnosis
of fracture of the femur, and 7 had other
fractures of the lower limb.

No difference in the number hospitalized was
demonstrated between nursed and non-nursed
subgroups of patients who had a muscular or
skeletal condition without a coexisting major
chronic condition, nor between those who had a
fractured femur without such comorbidity or
any lower limb fracture without comorbidity.
We did not demonstrate difference in physician
contacts between nursed and non-nursed pa-
tients who had a muscular or skeletal condition
with or without a coexisting major chronic
condition. Similarly, differences in physician
contacts were, similarly, not demonstrated for

Effects of Treatment Program

those who had a fractured femur or any lower
limb fracture with or without major comor-
bidity.

Patients in the nursed group who had a
principal diagnosis of stroke plus a coexisting
major chronic condition had more contacts with
physicians during the two weeks before their
terminal interview (15 of 24) than did similar
non-nursed patients (4 of 17). Related and
consistent findings were more contacts with
physicians for patients who had stroke or
chronic brain syndrome plus major comorbidity
and for patients who had any brain condition
plus major comorbidity in the presence of the
nursing program (15 of 28 with stroke or
chronic brain syndrome and 17 of 31 with any
brain condition) than in its absence (5 of 20
with stroke or chronic brain syndrome and 7 of
25 with any brain condition).

In comparing nursed and non-nursed sub-
groups, we did not demonstrate differences
between the numbers admitted into hospitals for
the following classes of principal diagnoses (with
or without major comorbidity): stroke, stroke
or chronic brain syndrome, and any brain
condition. Contacts with physicians did not
differ among patients in the following classes:
stroke without a coexisting major chronic condi-
tion, stroke or chronic brain syndrome without
a coexisting major chronic condition, and any
brain condition without a coexisting major
chronic condition.

Nursed patients with a principal diagnosis of
cardiovascular-renal disease plus a coexisting
major chronic condition had more frequent
contacts with physicians during the two weeks
before their terminal interview (19 of 30 pa-
tients) than did corresponding patients in the
non-nursed group (4 of 20 patients). The same
diagnostic class had more in-hospital deaths in
the nursed group (11 of 16 who died) than in
the non-nursed group (2 of 11 who died). No
differences in the numbers of patients hospital-
ized were demonstrated between nursed and
non-nursed patients who had cardiovascular-
renal disease with or without major comor-
bidity, and no difference was demonstrated
with respect to contacts with physicians for

65
Effects of Continued Care

those who had cardiovascular-renal disease with-
out major comorbidity.

Next, we consider outcome with respect to
disease-disability profiles. Among patients classi-
fied according to the total burden of illness in
terms of disease-disability profiles which were
described in Chapter 6, more of those with
maximum disability were hospitalized in the
presence of the nursing program than in its
absence. Here, 30 of 41 nursed patients were
hospitalized as compared with 20 of 39 non-
nursed patients. Those in the second severity
class (having less than maximum disability plus
major comorbidity) had fewer admissions to
nursing homes in the presence of the nursing
program (5 of 40 as compared with 15 of 46 in
the program’s absence). In the absence of both
maximum disability and major comorbidity,
more patients assigned to the nursing program
were hospitalized (26 of 36) than those not so
assigned (18 of 37); and those assigned to the
nursing program received more non-nurse and
non-physician professional services. In this case,
36 of 48 nursed patients received non-nurse
and non-physician professional services in con-
trast to 28 of 50 non-nursed patients.

In this and the following paragraph, we
present outcome with respect to disability,
mental function, and cardiovascular-renal dis-
ease. Of patients who were severely disabled in
terms of activities of daily living (grades E, F,
or G by the Index of ADL), 75 of 100 were
hospitalized in the presence of the nursing
program as opposed to 50 of 97 in its absence.
In the presence of the nursing program, patients
in this same dependent category had more
contacts with physicians during the two weeks
before their terminal interviews (40 of 75 as
compared to 27 of 75), had more non-nurse and
non-physician professional services during the
three months preceding the terminal interview
(48 of 73 as compared to 34 of 72), more
frequently died in a hospital (25 of 39 who died
as compared to 13 of 32), and had more
autopsies. Patients in the relatively less depend-
ent category of activities of daily living (grades
A, B, C, or D by the Index of ADL) had fewer
admissions to nursing homes in the presence of
the nursing program than in its absence. In this

66

subgroup, 1 of 27 nursed patients were admitted
to nursing homes as opposed to 9 of 36
non-nursed patients.

Poor mental function and cardiovascular-renal
abnormalities were among the important abnor-
malities included in the classification terms
“coexisting major chronic condition” and
“major comorbidity”. Poor mental function was

also represented in the major disability classes,
since poor mental function was often the basis

for major dependence in activities of daily living
and the two were strongly correlated with each
other. Poor mental function and cardiovascular-
renal abnormalities were, thus, factors which
contributed to the previously described signifi-
cant outcomes for patients with major comor-
bidity or major dependence in activities of daily
living. When we examined study results in
patients classified according to mental function
or cardiovascular-renal status without taking
other disease or disability factors into account,
we did not demonstrate differences in use of
services between patients assigned and not as-
signed to the nursing program.

In summary, the findings in this chapter
demonstrate that mildly or moderately disabled
patients showed physical and psychological
benefits in the presence of the nursing program,
as did those with uncomplicated musculoskeletal
disease and those who functioned adequately
mentally. The mildly or moderately disabled
group also had fewer admissions to nursing
homes. Severely ill and disabled patients, on the
other hand, did not show similar beneficial re-
sults, but made more use of hospital and other
professional health services, as did those with
complicated central nervous system or cardio-
vascular-renal disease. Decreased social interac-
tion was a nursing effect experienced by those
with poor mental function, and those with
severe cardiovascular-renal disease. The foregoing
were consistent and significant effects, not
biased by introducing observers separately or in
parallel with the nursing program (see summaries
of “observation” and “interaction” effects in
Chapter 11).

Discussions of nursing effects in both this
chapter and in Chapter 9 are based on the most
frequently encountered outcomes, where we
define the term “most frequently encountered”
as any outcome which was statistically signifi-
cant for at least five per cent of the study’s
Intake Classes (listed in Chapter 14). Significant
outcomes which occurred less frequently (gener-
ally for only one or two of the 71 Intake
Classes) were more likely to represent infrequent
random occurrences and are summarized separ-
ately as follows. Nine of the 16 infrequent
outcomes were in the same direction as fre-
quently occurring outcomes, namely, more
beneficial effects associated with the nursing

Effects of Treatment Program

program, decreased social interaction, and
greater use of services. The only infrequent
outcome which occurred in as many as three
Intake Classes was a beneficial nursing effect in
terms of fewer deteriorations by the Raven test
of observation and clear thinking. The descrip-
tions in Chapters 7 and 9 are consistent with 93
per cent of the study’s significant findings. The
descriptions are, thus, not changed materially by
information about infrequent outcomes and can
be accepted with a high degree of confidence.

67
CHAPTER VIII.

An essential quality of chronic illness is its
duration. The illness becomes a part of the
continuing personal and social life of the af-
flicted individual, and factors in the social
environment may contribute to the development
of the illness and influence its course. A longi-
tudinal study of chronically ill patients must,
therefore, take account of social factors in
describing the course of events and in evaluating
the results of treatment.

In a review of “‘social psychological factors in
illness”, King observes that conditions or events
in the social environment contribute to the
etiology of disease and to the course of dis-
ease (1). Here, the relationship ranges from a
clear-cut one, such as malnutrition resulting
from culturally-determined dietary habits, to an
indirect one such as the relationship between
tuberculosis mortality and environmental fac-
tors, namely, housing and income. Although
direct causal relationships between social factors
and disease, or the course of disease, have been
established in only a few instances, there are
many demonstrated associations with social fac-
tors which have yet to be explained. This being
the case, it is necessary for the investigator of a
chronic disease to identify and measure those
social characteristics which seem most relevant
to the research question. In addition to previ-
ously defined primary groupings, patients in our
study were, therefore, classified in terms of the
following social characteristics as they entered
the study: (1) identifying social characteristics
(including social class, economic dependence,
social role, and household composition), (2) so-
cial interaction, and (3) social deprivation. We
present the description of the social categories in
this chapter.

Identifying Social Characteristics at Intake
Relationships between health and social factors

68

CLASSIFICATION OF SOCIAL FACTORS FOR THE STUDY

such as social class and economic status have
been demonstrated in several fields as, for
example, maternal and child health, general
health and mortality, and chronic illness
(2, 3,4,5). The study of social class and mental
illness by Hollingshead and Redlich is of particu-
lar interest, since it demonstrates that treatment
and outcome in comparable illness situations
may also be correlated with social position (6).

We used several measures to define social
characteristics of patients at the time they
entered the study. As one measure, we selected
Hollingshead’s “Two Factor Index of Social
Position” based on occupation and educa-
tion (7). None of the available indices, including
Hollingshead’s, is well standardized for the kinds
of elderly people seen in this study, many of
whom represent fragments of the nuclear family
which forms the basis of classification for many
social indices. For this reason, we also used the
“Index of Economic Dependence”, which is a
modification of a measure developed in the
course of studying similar patients discharged
from the same hospital (8). The measure is
described in Chapter 5 and is based on the
following three characteristics: (1) Financial
support from a charitable agency, (2) employ-
ment status, and (3) home ownership. These
economic factors are clearly both indicators of
the financial situation in which a particular
patient’s illness occurs and measures of certain
resources available for care. Classification of
patients at intake was expressed in terms of the
four classes of the index, Class 1 being economi-
cally the most independent and Class 4 the most
dependent.

Social role or function may theoretically be
distinguished from status or position. The roles
of a man as worker and of a woman as
homemaker are probably the most critical social
functions for the elderly. Retirement from
either role alters an individual’s socioeconomic
function immediately and may eventually alter
status through loss of income or increased
expenses. Because of the importance of these
roles in a group such as the present one, those
who were actively working or homemaking until
the illness for which they were hospitalized were
identified as groups for separate study.

We classified household composition in terms
of three categories; married and living with
spouse, solo living with others, and living alone.
“Solo” means unmarried, widowed, separated,
or divorced. Marital status was included in the
definition of household composition; since, for a
sick or disabled person, his relationship to those
with whom he lives strongly influences how
much assistance he can expect from them. For
this reason, also, we defined a separate category
as “living with spouse or child”.

Social Interaction at Intake

Man is a social animal, and every individual
participates in many kinds of formal and in-
formal contacts with other human beings. The
number and variety of these contacts tend to
diminish with age and illness. A few individuals
choose lives of social isolation while more have
it thrust on them.

Associations between illness and social isola-
tion have been reported in such studies of the
aged as those of Kutner and of Townsend (5, 9).
In another report, mortality rates have been
found to be lower for the married than for the
single or widowed (10). The bases for these
relationships between social status and health
status are not well-established. Interpretation of
the findings is hampered by difficulties in
designing and conducting causal studies, as well
as by problems of measuring social interaction.
Terminology differs from one investigator to the
next. For example, Cumming and Henry speak
of “engagement” and use a “Social Lifespace”
measure based on weighted scores for contacts
with relatives, friends, neighbors, and
others (11). Townsend uses a ‘“‘scale of isola-
tion” derived from the same kind of data (9).
Kutner refers to ‘“‘range of interpersonal rela-
tions’ or “relative social isolation” based on
broadly defined kinds of social interaction (5).

Social Factors for the Study

Based on the above ideas and in the absence
of any standard measure, we collected a large
number of observations bearing on social inter-
action (Forms A-2, A-3, A-4, and A-5 in Chapter
13). As the study progressed, we organized these
data for purposes of analysis into a series of
measures. Classes embodied in the measures
were defined with the requirement in mind that
they had to contain large enough numbers of
patients to permit comparison between experi-
mental groups.

The first measure was a scale of personal
interaction, based on the number of kinds of
individuals with whom the subject had daily
contact. The kinds of individuals were classed as
spouse, relatives, or friends; and the three grades
of the scale were defined as daily contact with
two or three categories of persons, with one
category, and with none. This measure included
contacts with people both within and outside of
the household.

Three separate indicators were designated as
measures of social activity or inactivity. These
were “talks to more than five persons daily”,
“any regular organized social activities”, and
“employed during the two weeks prior to intake
hospitalization’. Finally, we synthesized an
additional three indicators of interaction from
the above elements of social interaction, namely,
from elements of personal interaction and social
activity.

Social Deprivation at Intake

Parkes reports that the loss of a husband
increases the frequency with which widows
consult their physicians for non-psychiatric
symptoms (12). Townsend describes circum-
stances of subjectively perceived personal loss or
“desolation” in older people and reports an
increased death rate among the older widowed
population, while Lowenthal and Boler describe
a correlation between low morale and recent
retirement, widowhood, and physical disability
(9, 13). “Social deprivation” was the term we
used to refer to this recognized, but not clearly
defined, special form of isolation which is
characterized by recent social losses of older
people and which may be a precursor of loneli-
ness (14).

69
Effects of Continued Care

The measures of social deprivation in our
study included: ‘lost spouse in the two-year
period prior to study”, “retired within two years
prior to admission to Abington House”, and
“ceased homemaking within two years prior to
admission to Abington House”. We also defined
combined categories of the above measures of
social loss. A rough measure related to the
Cumming and Henry concept of “Perceived Life
Space” was included to obtain more information
about the impact of social loss on study results
(11). This measure was “talks to fewer persons
at intake than at age 45”.

References

1. King, S. H.: Social Psychological Factors in Illness,
Handbook of Medical Sociology (ed. Freeman, H.
E., Levine, S. and Reeder, L. G.) Published by
Prentice Hall, Inc., Englewood, N. J., pp. 99-121,
1963.

2. Boek, W. E., Lawson, E. D., Yankauer, A. and
Sussman, M. B.: Social Class, Maternal Health and
Child Care. Published by the New York State
Department of Health, Albany, N. Y., pp. 65-118,
1958.

3. Stockwell, E. G.: Use of Socioeconomic Status as a
Demographic Variable. Public Health Reports 81:
961-966, Nov. 1966.

4. Commission on Chronic Illness: Chronic Illness in
the United States. Vol. 1V, Chronic Illness in a
Large City. Published for the Commonwealth Fund,
Harvard University Press, Cambridge, Mass., pp.
151-170, 1957.

70

10.

11.

12.

13.

14.

. Kutner, B., Faushel, D., Togo, A. M. and Langner,

T. S.: Five Hundred Over Sixty: A Community
Survey on Aging. Published by the Russell Sage
Foundation, N. Y., N. Y., pp. 137-147, 1956.

. Hollingshead, A. B. and Redlich, F. C.: Social Class

and Mental Illness. Published by John Wiley and
Sons, Inc., N. Y.,N. Y., pp. 253-303, 1958.

. Hollingshead, A. B.: Two Factor Index of Social

Position. Published by Hollingshead, New Haven,
Conn., pp. 1-11, 1957.

. The Staff of The Benjamin Rose Hospital: Multidis-

ciplinary studies of illness in aged persons. V. A new
classification of socioeconomic functioning of the
aged. J. Chron, Dis. 13: 453-464, 1961.

. Townsend, P.: The Family Life of Old People.

Published by Routledge and Kegan Paul, Ltd.,
London, England, pp. 178-182, 1957.

Shurtleff, D.: Mortality and Marital Status. Public
Health Reports, 70: 248-252, March 1955.

Cumming, E. and Henry, W. E.: Growing Old: The
Process of Disengagement. Published by Basic
Books, Inc., N.Y., N.Y, pp. 47-51 and 244-250,
1961.

Parkes, C. M.: Effects of Bereavement on Physical
and Mental Health—A Study of the Medical Records
of Widows. Brit. Med. J. 2: 274-279, 1964.

Lowenthal, M. F. and Boler, D.: Voluntary vs.
Involuntary Social Withdrawal. J. Gerontology 20:
363-371, 1965.

Townsend, P.: Old People in Three Industrial
Societies. Published by Atherton Press, N. Y.,N.Y.,
pp. 275-286, 1968.
CHAPTER IX. EFFECTS OF TREATMENT PROGRAM: SUBGROUPS
BY AGE, SEX AND SOCIAL CHARACTERISTICS

In the preceding Chapter, we presented the
definitions and rationale for the social groupings
of patients. In this chapter, we report significant
associations between the treatment program and
the outcome for subgroups of patients which
were similar in age, sex, identifying social
characteristics, social interaction, and social
deprivation as they entered the study. In the
first section of the chapter, we discuss effects of
the nursing program in terms of function (physi-
cal, mental, and social), injury, and mortality.
The second section of the chapter deals with the
outcome in terms of use of services.

In relation to physical and mental function,
evaluations of outcome were limited to systems
of classification which identified improvement
or maintenance (not improvement alone), since
the study’s main hypothesis was that function
would be improved or maintained in the pres-
ence of the experimental care variable. As we
discussed in Chapter 7, definitions of change in
physical, mental, and social function excluded
death as an item of change.

Outcome in Terms of Function,
Injury, and Mortality

Among certain subgroups which were homoge-
neous for age, sex, and social characteristics, statis-
tically significant and consistent differences in
function were demonstrated between patients
who were assigned to the nursing program
and those who were not. The most frequent
outcomes in the presence of the nursing program
were: (a) beneficial nursing effects in physical
and/or mental function among the youngest
(50-64 years), those who had been employed or
homemaking before the present disabling illness,
those who had not lived alone, those who had
been socially active, and those who had been
socially nondeprived; (b) increased house-confine-
ment among men, those who were relatively
independent economically, those who had lived
with a spouse, and those who had been socially
active; and (c) decreased social interaction among

the oldest, those who were relatively dependent
economically, and those who reported a prior
narrowing of social contacts.

The first groups of patients we shall discuss
are those homogeneous in terms of age and sex.
Patients who were 50 through 64 years old at
the time they entered the study (the youngest
patients) and who were assigned to the nursing
program had fewer deteriorations in range of
movement and fewer deteriorations in orienta-
tion and mental control than did similar patients
not assigned to the nursing program (Table 9.1).
For patients aged 65 through 74 years, the
presence of the nursing program did not have
any significant effect on function, injury, or
mortality, while nursed patients who were 75
years old or older were more likely to decrease
their participation in organized social activities
than respective patients in the non-nursed group
(15 of 35 patients in the nursed group and 6 of
43 in the non-nursed group). Concerning effects
according to sex, increased house-confinement
occurred more frequently among men in the
nursed group (14 of 34) than among men in the
non-nursed group (5 of 28).

We consider, next, outcome in patients classi-
fied according to their identifying social charac-
teristics at the time they entered the study.
Among patients who were relatively independ-
ent economically and patients who were married
and living with a spouse, increased house-
confinement was more frequent in the groups
assigned to the nursing program than in the
respective groups not so assigned. Of the 51 in
the nursed group who were relatively indepen-
dent economically, 16 became more confined
to their homes in contrast to 6 of 47 in the
non-nursed group. Among those who lived with
their spouses, the respective numbers were 18 of
47 in the nursed group and 2 of 30 in the
non-nursed group. Patients who had been

71
Effects of Continued Care

Table 9.1 Nursing Program Effect on Physical and
Mental Function Two Years After Study Intake
in Patients 50 Through 64 Years Old

 

 

 

Patients 50 through
64 Years Old!
Referred Not Totals
Cs Referred
to Visiting to Visiting
Nurse (N+) Nurse (N-)
Range of Movement
(upper limb)
Deterioration? 0 6 6
No deterioration 31 24 55
Range of Movement
(any limb)
Deterioration? 2 9 11
No deterioration 29 21 50
Orientation and
Mental Control
Deterioration? 6 13 19
No deterioration 22 15 37

 

 

 

 

 

 

L Ages are the ages of patients as they entered the study.

2Deterioration in range of movement (upper limb) is Outcome
Class #7 defined in Chapter 14. Deterioration in range of move-
ment (any limb) is Outcome Class #9. Deterioration in orien-
tation and mental control is Outcome Class #11.

employed or homemaking until they were hospi-
talized in Abington House had fewer deteriora-
tions in the presence of the nursing program (3
of 35 patients) by the test of range of movement
than in the absence of the program (14 of 50
patients). Fewer deteriorations by the test of
orientation and mental control occurred among
nursed patients who lived with people other
than spouses (5 of 27 patients) than among
similar non-nursed patients (15 of 35 patients).
Patients who were relatively dependent eco-
nomically and in the nursed group decreased
social interaction more frequently (19 out of 48
patients) than did similar patients who were in
the non-nursed group (12 of 59 patients), when
the measure of social interaction was defined as
the change in the number of organized social
activities between the intake and terminal points
of the study.

72

Classification of patients in terms of social
interaction at the time they entered the study
resulted in subgroups of patients who had been
relatively active socially before the present
illness, relatively inactive, or isolated. The most
active subgroup had fewer deteriorations in
range of movement in the presence of the
nursing program than in its absence (Table 9.2).
We defined a slightly less active class of patients,
simply, as those who had had daily contacts
with at least two of three categories of people
(spouse, relatives, friends). The nursed patients
in this subgroup showed more frequent increases
in the degree of house-confinement (11 of 28)
than did corresponding patients in the non-
nursed subgroup (2 of 19).

As discussed in Chapter 8, we also classified
patients as socially deprived or socially nonde-
prived. In comparisons of outcome, nursed
patients in one of the socially deprived sub-
groups (specifically, those who had reported a
narrowing of social contacts since age 45) more
frequently decreased their participation in social
activities in the presence of the nursing program

Table 9.2—Nursing Program Effect on Range of
Movement Two Years After Study Intake in
Socially Active Patients

 

 

 

Socially Active
Patients!
Referred Not Totals
Referred
to Visiting to Visiting
Nurse (N+) Nurse (N-)
Range of Movement
(upper limb)
Deterioration? 0 5 S
No deterioration 23 9 32

 

 

 

 

 

 

Lor this table, socially active patients are those who had daily
contact with at least two of three categories of people (spouse,
relatives, and friends) and who also either had talked to more
than five people per day, had participated regularly in social
activities, or had been working immediately before hospitaliza-
tion (Intake Class #65 defined in Chapter 14).

Deterioration in range of movement (upper limb) is Outcome
Class #7 defined in Chapter 14.
(25 of 67 patients) than in its absence (12 of 77
patients). Patients who were socially nonde-
prived had fewer deteriorations in range of
movement (Table 9.3).

Outcome in Terms of Use of Services

As we described in Chapter 7, hospitalizations
and autopsies were more frequent among the
150 patients assigned to the nursing program
than among the 150 not so assigned. This
section deals with differences in use of services
between subgroups which, at the time of entry
into the study, were relatively homogeneous
according to age, sex, and social characteristics.
(As was naturally expected, more visiting nurse
visits were received by all subgroups of patients
who were designed to receive such service).
Among subgroups which were homogeneous for
age, sex, and social characteristics at the time of
entry into the study, the most frequent outcomes
in the presence of the nursing program were:
(a) more patients hospitalized among the oldest,
those who had been socially active before the
present disabling illness, and those who had
recently been socially deprived; (b) more contacts
with physicians among men, those in higher social
classes, those who lived with a spouse, and those

Table 9.3—Nursing Program Effect on Range of
Movement Two Years After Study Intake in
Socially Nondeprived Patients

 

 

 

Socially Nonde-
prived Patients!
Referred Not Totals
ben Referred
to Visiting to Visiting
Nurse (N+) Nurse (N-)
Range of Movement
(any limb)
Deterioration? 10 21 31
No deterioration 56 49 105

 

 

 

 

 

 

Socially nondeprived patients are those who had not lost their
spouse in the 2-year period prior to the study or who had not
terminated employment or homemaking between 30 days and
two years before admission to Abington House (Intake Class
#69 defined in Chapter 14).

2Deterioration in range of movement (any limb) is Outcome
Class #9 defined in Chapter 14.

Effects of Treatment Program

who had been socially active before the present
disabling illness; (c) more non-nurse and non-
physician professional services among those who
were relatively independent economically, those
who did not live alone, and those who had been
relatively active socially; (d) fewer nursing home
admissions among those who had recently been
socially deprived; and (e) more hospital services at
the time of death among those who had been
economically dependent, had been inactive so-
cially, and had recently been socially deprived.

There was only one subgroup of patients
grouped according to age for which we demon-
strated a significant difference in use of services
between those who were and were not assigned
to the nursing program: more patients who were
75 years old or older were hospitalized in the
presence of the nursing program than in its
absence (Table 9.4). In this aged group, 72 per
cent of those assigned to the nursing program
were hospitalized (39 of 54) in contrast to 49
per cent of those who were not assigned to the
nursing program (29 of 59).

The one difference which we demonstrated in
nursed and non-nursed subgroups categorized
according to sex was that significantly more men
in the nursed group had contacts with physicians
during the two weeks before their final study
evaluations than men in the non-nursed group.
Among men, 59 per cent of the 34 referred to
the nursing program for the study had contacts
with physicians, while the proportion was 29 per

Table 9.4—Nursing Program Effect on Hospitalization
During the Study in Patients 75 Years Old or Older

 

 

 

 

 

 

 

Patients 75 Years
Old or Older!
Referred Not Totals
Cs Referred
to Visiting to Visiting
Nurse (N+) Nurse (N-)
Hospitalized
Yes? 39 29 68
No 15 30 45

 

 

1 Ages are the ages of patients as they entered the study.
2Hospitalized during the study is the Outcome Class #19 defined
in Chapter 14.

73
Effects of Continued Care

cent of the 28 who were not referred to the
nursing program.

Considering patients grouped according to
identifying social characteristics, those in rela-
tively high social classes (Hollingshead’s Index
classes I, II, and III) and those married and living
with spouse more often had contacts with
physicians during the two weeks before their
final study evaluations in the presence than in
the absence of the nursing program. Those
contacting physicians in the nursed group com-
prised 57 per cent of the 42 in high social classes
(as opposed to 36 per cent of 45 in the
non-nursed group) and 57 per cent of 47
married and living with spouse (as opposed to 32
per cent of 31 in the non-nursed group). During
the three months preceding the terminal evalu-
ations, the following nursed social subgroups
received more non-nurse and non-physician pro-
fessional services than did similar non-nursed
subgroups: patients who were relatively inde-
pendent economically (32 of 50 patients in the
nursed group as compared to 18 of 45 patients
in the non-nursed group) and patients who did
not live alone (23 of 28 patients in the nursed
group as compared to 20 of 37 patients in the
non-nursed group). More autopsies were
performed in the presence of the nursing pro-
gram on deceased patients who had been eco-
nomically dependent than in the absence of the
nursing program (11 of 26 in the nursed group
as opposed to 3 of 22 in the non-nursed group).

We discuss, next, outcome among patients
classified for social interaction. Patients who had
been socially active before their present dis-
abling illness were more likely to be hospitalized
in the presence than in the absence of the
nursing program (Table 9.5). During the two
weeks before the final study evaluations, signifi-
cantly more contacts with physicians were made
by nursed (18 of 28) than non-nursed (6 of 20)
patients who had daily contacts with at least
two of three categories of people (spouse,
relatives, and friends). The subgroup of patients
who were intermediately active (interacted daily
with either spouse, relatives, or friends) received
more non-nurse and non-physician professional
services during the three months preceding the
terminal evaluation in the presence than absence

74

Table 9.5—Nursing Program Effect on Hospitalization
During the Study in Socially Active Patients

 

 

 

Socially Active
Patients!
Referred R he d Totals
to Visiting Vit
Nurse (N+) to Visiting
Nurse (N-)
Hospitalized

Yes? 36 22 58
No 16 23 39

 

 

 

 

 

 

Igor this table, socially active patients are those who had daily
contact with one of three categories of people (spouse, rela-
tives, and friends) and who also either had talked to more than
five people per day, had participated regularly in social activi-
ties, or had been working immediately before hospitalization
(Intake Class #66 defined in Chapter 14).

2Hospitalized during the study is the Outcome Class #19 defined

in Chapter 14.

of the nursing program (39 of 58 patients as
compared to 25 of 58 patients). Nursed patients
who had participated regularly in social activities
before entering the study were more often in
hospitals at the time of death than similar
non-nursed patients (15 of 20 patients as op-
posed to 5 of 14). More autopsies were per-
formed at the time of death on deceased
patients in the following nursed subgroup than
on similar non-nursed patients: those who had
daily contact with no one (7 of 10 in the nursed
group as opposed to 0 of 5 in the non-nursed
group).

Outcomes related to social deprivation are
considered in this paragraph. Patients who had
been assigned to the nursing program and who
had suffered the deprivation of a recent loss of
spouse or cessation of employment or home-
making were more likely to be hospitalized
(Table 9.6) and had fewer admissions to nursing
homes than similar patients who had not been
assigned to the nursing program. In this class of
patients, the respective numbers of admissions
to nursing homes were 4 of 37 patients assigned
to the nursing program as compared with 15 of
Table 9.6—Nursing Program Effect on Hospitalization
During the Study in Socially Deprived Patients

 

 

 

Effects of Treatment Program

Table 9.7—Nursing Program Effect on Hospitalization
During the Study in Patients with
Narrowed Social Contacts

 

 

 

 

 

 

Socially Deprived
Patients!
Referred Not Totals
es Referred
to Visiting to Visiting
Nurse (N+) Nurse (N-)
Hospitalized
Yes? 35 27 62
No 12 22 34

 

 

 

Patients
with Narrowed
Social Contacts!
Referred Not Totals
Cs Referred
to Visiting to Visiting
Nurse (N+) Nurse (N-)
Hospitalized
Yes? 74 54 128
No 28 48 76

Socially deprived patients are those who had lost their spouse
in the 2-year period prior to the study or who had terminated
employment or homemaking between 30 days and two years
before admission to Abington House (Intake Class #68 defined
in Chapter 14).

2Hospitalized during the study is the Outcome Class #19 defined
in Chapter 14.

44 patients not assigned to the nursing program.
More autopsies were performed at the time of
death on deceased patients of the nursed group
who had recently retired or ceased homemaking
than on similar non-nursed patients (7 of 12 as
opposed to 1 of 11). Finally, we demonstrated
the following significant differences between
nursed and non-nursed patients who had nar-
rowed their social contacts between age 45 and
the intake point of the study: more patients
hospitalized (Table 9.7), more in-hospital deaths
and more autopsies at the time of death in the
presence than absence of the nursing program (8
of 15 opposed to 1 of 14).

In summary, younger patients showed physi-
cal and psychological benefits in the presence of
the nursing program, as did those who had not
lived alone before the present disabling illness
and those who had been employed, home-
making, or otherwise active socially. Decreased
social interaction was a nursing effect experi-

 

 

 

 

 

 

Ipatients with narrowed social contacts are those who talked to
fewer people than they had at age 45 years (Intake Class #70
defined in Chapter 14).

2Hospitalized during the study is the Outcome Class #19 defined

in Chapter 14.

enced mainly by the oldest, recently socially
deprived, and economically dependent patients.
Patient subgroups which showed decreased
social interaction were generally the same as the
subgroups which had more hospitalizations and
more hospital services at the time of death.
Increased house-confinement was an effect for
men living with a spouse, for those who had
been socially active, and for those who were
economically independent. Patient subgroups
which showed increased house-confinement
tended to be the same as the subgroups which
used more physician and other professional
services. Among those who had recently been
socially deprived, fewer nursing home admis-
sions occurred in the presence of the nursing
program. The foregoing were consistent and
significant effects, not biased by introducing
observers separately or in parallel with the
nursing program (see summaries of ‘“‘observa-
tion” and “interaction” effects in Chapter 11).

75
CHAPTER X.

Chapters 7 and 9 dealt with treatment effects
at the end of two years. In this chapter, we
describe the times of appearance and durations
of the treatment effects during the two year
period of the study. Dates of death, injuries,
admissions to nursing homes, and admissions to
hospitals were identified for all 300 patients;
thus, treatment effects with respect to these
outcomes could be determined exactly and,
according to the study design discussed in
Chapter 2, without the bias of the observation
process. With respect to physical, social, and
economic functioning, as well as use of profes-
sional services, we identified treatment effects at
scheduled times during the study, namely, semi-
annually for social functioning and quarterly for
the remaining types of outcomes. These out-
comes could only be evaluated for the 75
treatment and 75 control patients who were
scheduled for quarterly observations, and the
evaluations were not available for the 150
patients who were not so scheduled. As dis-
cussed in Chapter 2, such treatment effects
(determined only in quarterly observations)
were subject to a possible observer bias or to a
possible interaction bias (simultaneous presence
of the observation process and treatment pro-
gram). We describe interval results in the present
chapter, therefore, under two headings, “Un-
biased Interval Results” and “Interval Results
Subject to Possible Bias”.

Consistent with the approach used through-
out the book to describe effects of the nursing
program, discussions in this chapter are based on
the most frequently encountered outcomes,
where we define the term ‘“‘most frequently
encountered’ as any outcome which was statisti-
cally significant for at least five per cent of the
Intake Classes which were evaluated. As we
described in Chapter 9, the great majority of
infrequent outcomes were in the same direction

76

INTERVAL RESULTS

as frequently occurring ones and did not change
interpretations concerning study results. Finally,
a brief comment about the term “interaction”
will help to avoid possible confusion arising
from two uses of the term. The term “inter-
action bias” or “‘interaction effect’ applies to an
effect measured by the study design as described
in Chapter 2, while the term “social interaction”
refers to an area of social function.

Unbiased Interval Results

Nursing program effects appeared six to twelve
months after the treatment program was intro-
duced. Starting at that time, fewer patients were
admitted to nursing homes in the presence of the
treatment program, and more were hospitalized.
Survival did not differ significantly between treat-
ment and control groups; and although fractures
tended to appear earlier in the treated group, the
actual number of such injuries did not differ
between the two groups.

As previously stated, we could determine
interval effects in terms of death, injuries,
admissions to nursing homes, and admissions to
hospitals exactly and without the bias of the
observation process. During the course of the
study, 29 per cent of the 300 patients died: 12
per cent sustained fractures: 20 per cent were
admitted into nursing homes: and 57 per cent
were hospitalized.

With respect to mortality, we did not demon-
strate differences between those assigned to the
nursing program and those not so assigned. In
Figure X-1, we compare survival in nursed and
non-nursed groups in graphs made with the aid
of life table methods (1).

Nine months after the treatment program was
introduced, injuries in terms of fractures ap-
peared in certain subgroups more frequently in
the presence of the nursing program than in its
absence. This difference persisted for the next
nine months, after which time we demonstrated
 

Interval Results

   

 

1.00
90
80
s Tor O Referred to visiting nurse (N+)
B60F A Not referred to visiting nurse (N-)
2 |
© 50
= _
S
§ 40r
oO be
& 3of
20}
10}
0.00 I 1 1 1 1 1 1 1 1
0 3 6 9 4 15 18 2 24

TIME SINCE ADMISSION INTO STUDY (MONTHS)

Figure X-1. Survival in Patients Referred and Not Referred to Nursing Program

no difference between treatment and control
patients. Since the occurrence of fractures was
analyzed as a cumulative event over time,
disappearance of the difference between the
treatment and non-treatment groups meant that
there was no final difference in the actual
number of fractures during the study; however,
such injuries appeared earlier in the presence of
the treatment program (Figure X-2). We noted
the earlier appearance of injuries among people
who had been employed or homemakers before
the present disabling illness, among people who
had lived alone and had not been socially active,
and among those who were markedly disabled as
they entered the study.

Six months after we introduced the treatment
program, a treatment effect on nursing home
admissions became apparent in certain sub-
groups (Figure X-3). At that time, fewer patients
were admitted to nursing homes in the presence
of the nursing program than in its absence; and,

with the exception of one 3-month period, this
finding persisted in a significant number of
subgroups during the final 18 months of the
study. The number of subgroups which showed
fewer nursing home admissions during successive
3-month periods varied from three to nine; and,
with few exceptions, these subgroups were the
same as or similar to the subgroups in which the
nursing home effect was present at the end of
the study. In Chapters 7 and 9, we described the
subgroups with fewer nursing home admissions
at the end of the study.

They included patients who were not maxi-
mally disabled, patients who were not maxi-
mally disabled and had indicators of major
chronic conditions, and patients who had been
recently deprived socially (widowed, retired, or
ceased homemaking). Additional subgroups with
fewer admissions during the course of the study
included those who were least ill, those who
lived with others, and those who were economi-
cally relatively dependent. Associated with the

717
Effects of Continued Care

100

90
80
TOF
sof
SOF
40f
30

PROPORTION WITH FRACTURE

20
A0F

fore BE oN

0.00!
o

— >

 

b w—ce

3 6 9

- O Referred to visiting nurse (N+)
A Not referred to visiting nurse (N-)

 

a

24

_o—0

ee \p——\
or TT

12 15 8 2

—O0—
ed)
1

TIME SINCE ADMISSION INTO STUDY (MONTHS)

Figure X-2. Fractures in Patients Referred and Not Referred to Nursing Program

finding of fewer nursing home admissions, the
percentage of days spent in nursing homes was
consistently lower in the treatment group
(Table 10.1).

Six months after the treatment program was
introduced, a treatment effect on hospitaliza-
tions appeared, namely, more patients hospital-
ized in the presence of the nursing program than
in its absence. Figure X44 illustrates graphically
the proportions of patients readmitted to hos-
pitals after discharge from Abington House
throughout the 2-year period of the study. This
effect persisted through the end of the study,
and the number of subgroups in which the
finding was present during subsequent 3-month
periods varied from 12 to 21. In Chapter 7, we
observed that the effect with respect to hospital-
izations applied to the total group of patients, as
well as to subgroups. The hospitalization effect
was strong. In addition to the fact that the
effect for the total group was present in every
quarterly interval after the first half year, 44 per

78

cent of the 71 subgroups showed this effect at
sometime during the course of the study. The
following were the prevailing subgroups: pa-
tients who had principal diagnoses of fractured
femur with indicators of coexisting major
chronic conditions, patients who had principal
diagnoses of any lower extremity fracture with
indicators of coexisting major chronic condi-
tions, the most disabled patients, those without
both maximum disability and indicators of
major chronic conditions, women, patients aged
75 years or older, those in low socioeconomic
classes, those who were relatively dependent
economically, those who lived with others, those
who had been moderately active socially before
the present disabling illness, patients who talked
with fewer people than they had at age 45, and
patients who had not shown recent social
deprivation (widowed, retired, or ceased home-
making). Though the number of patients who
were hospitalized differed significantly between
treatment and control groups, the percentage of
 

  

Interval Results

    

   

 

 

1.00
wi pe
3 oof
I -
e 80
5 -  O Referred to visiting nurse (N+)
> Tor A Not referred to visiting nurse (N-)
a 60
(@) —
—
ao 20r
HP pe
= 401
S he
2 30}
> _
© 20+
E -
e .10}
&
a 0.00 1 1 1 1 i 1 |
0 3 6 9 12 1S 18 21 24

TIME SINCE ADMISSION INTO STUDY (MONTHS)

Figure X-3. Admission to Nursing Homes in Patients Referred and Not Referred to Nursing Program

days spent in hospitals was not notably different
(Table 10.1).
Interval Results Subject to Possible Bias

The presence of observers did not seem to bias
nursing program effects, but interaction between
the observation process and nursing program
tended to obscure some true nursing program
effects during the intervals. Treatment effects
which were not obscured and which were detected
both during and at the end of the study included
increased house-confinement, decreased participa-
tion in regular social activities, and more services
from professionals other than physicians or nurses.

As we described earlier, conclusions about the
true nursing program effects required measure-
ments of the observer and interaction biases
(defined in Chapter 2); and though this was not
possible during the course of the study, the
study design permitted such measurements at
the terminal point of the study. By examining
interval outcomes in the light of observer and
interaction effects at the end of the study, we

could make limited interpretations about true
nursing effects during the intervals on a defined
and rational basis.

Interval evaluations were made of physical,
social, and economic functioning, and also of
use of professional services. On this basis, we
demonstrated major interval effects in terms of
the numbers of patients who became more
house-confined, who showed decreased social
interaction, who returned to employment, who
showed increased economic dependence, who
received more home visits from physicians, who
had more contacts with physicians, and who
received more services from other professionals.

Increased house-confinement and decreased
participation in regular social activities occurred
among patients who were assigned to both the
quarterly observations and the nursing program,
as compared to those assigned to quarterly
observations and not to the nursing program.
Increased house-confinement was intermittently

79
Effects of Continued Care

1.00
90
80

70r

 

—-  O Referred to visiting nurse (N+)
A Not referred to visiting nurse (N-)

 

 

J

=

a

wn

Oo

I

2

tw

= 50

Ss

2 a0

<2

oO 30

&

© .20

oO

T 10

000 1 1 1 1 | | 1 |
0 3 6 9 12 15 18 rl 24

TIME SINCE ADMISSION INTO STUDY (MONTHS)

Figure X-4. Admissions to Hospitals in Patients Referred and Not Referred to Nursing Program

present from the third quarter through the end
of the study, while decreased participation in
social activities was intermittently present from
the fourth quarter through the end of the study.
Both effects were major treatment effects at the
end of the study. Neither was an outcome
associated with the observation process at the
end of the study, and neither was an outcome
associated with the presence of an interaction
between the observation process and the treat-
ment program. We could, thus, reasonably in-
terpret the interval house-confinement and so-
cial participation effects as true nursing program
effects. Increased house-confinement was an
effect for the total group of observed patients
and for certain subgroups. Included in the
subgroups were patients who were most dis-
abled, patients without both maximum disabil-
ity and indicators of major chronic conditions,
men, those who lived with a spouse, those who
had been gainfully employed before the present
disabling illness, those who were economically

80

relatively independent, those in high socioeco-
nomic classes, and those who had been active
socially before their present illness. Subgroups
which demonstrated decreased participation in
social activities as a nursing effect included
patients who were not maximally disabled and
had indicators of major chronic conditions,
patients with major cardiovascular-renal condi-
tions, patients with poor mental function,
women, the oldest patients (75 years old or
older), those who were economically relatively
dependent, and those who talked to fewer
people than they had at age 45.

We observed two socioeconomic outcomes
during the course of the study which could not
be interpreted as nursing program effects,
namely, fewer returns to gainful employment
and more patients with increased economic
dependence, among those assigned to the quart-
terly observation process and nursing program in
contrast to those assigned to quarterly observa-
tions alone. In analyses at the terminal point of
 

Interval Results

Table 10.1 Locations of Patients in Nursed and Non-Nursed Groups During the Study

 

 

 

 

' Referred to Not referred to
Study interval Visiting Nurse (N+) Visiting Nurse (N-)
(months) Hospital Nursing Deceased Hospital Nursing Deceased
SP home or other p home or other
(percentage of person-days!)
0-3 10 1 89 10 1 89
3-6 6 3 91 3 6 91
6-9 4 5 91 3 6 91
9-12 4 5 91 4 9 87
12-15 4 8 88 3 11 86
15-18 3 7 90 3 10 87
18-21 2 8 90 5 11 84
21-24 4 9 87 4 13 83
Combined 24
months of study 4 6 20 4 8 88

 

 

 

 

 

 

 

 

percentages are the proportions of time spent by the entire group of patients in a given location. For a 3-month interval of 91
days, a group of 150 patients had a total of 13,650 (91 X 150) person-days of experience which were subdivided into percent-

ages of time in the various locations. For the combined 24-month period of 730 days, a group of 150 patients had a total of
109,500 (730 X 150) person-days of experience which were subdivided into percentages of time in the various locations.

the study, these socioeconomic outcomes were
neither nursing nor observer effects. They were
interaction effects; therefore, as interval find-
ings, they were probably, also, interaction ef-
fects. Confirmatory evidence was the fact that
the subgroups which showed these interaction
effects at the end of the study were similar to
the subgroups which showed these outcomes
during the intervals.

During the course of the study, patients
assigned to the nursing program and quarterly
observation process used more professional serv-
ices than those assigned to quarterly observa-
tions alone. The increased use of professional
services included more contacts with physicians,
more home visits by physicians, and more
services from other professionals. Two of these,
namely, increased contacts with physicians and
increased home visits by physicians, could not
be interpreted as nursing program effects. In
analyses at the terminal point of the study,
where the various types of study effects were
identified, increased contact with physicians was
not an observer effect. It was both a nursing

program effect and an interaction effect. Sub-
groups which showed this outcome as a nursing
program effect at the end of the study were
unlike subgroups which showed the interval
effect, while subgroups which showed the out-
come as an interaction effect at the end of the
study were similar to subgroups which also
showed the interval effect. On this basis, we
more reasonably interpreted the interval obser-
vation of increased contacts with physicians as
an interaction effect. Increased home visits by
physicians was, at the terminal point of the
study, neither a nursing nor an observer effect.
It was an interaction effect and so interpreted
for interval observations, again confirmed by the
similarity between subgroups which showed this
effect during and at the end of the study.

Among patients assigned to the group which
received both the nursing program and quarterly
observations, increased use of professional serv-
ices other than those of physicians and nurses
appeared twice during the course of the study.
This outcome appeared immediately after the
treatment program was introduced (within the

81
Effects of Continued Care

first week after discharge from Abington House)
and one year after the treatment program was
introduced. Since the nursing program was
barely in effect at the earlier time, we could not
attribute the increased use of services from
professionals, at that point, to the nursing
program. It was more likely to be attributable to
another factor such as the assignment process in
the study, which involved several days of close
work with patients’ families in preparation for
the introduction of the nursing program. At the
later time, however, we could reasonably con-
sider the increased use of other professional
services as a nursing program effect. For this
outcome, analyses at the terminal point of the
study revealed a nursing program effect and did
not reveal observer and interaction effects.
Similar subgroups were, furthermore, involved
during and at the end of the study. Included in
the subgroups were patients without both maxi-

82

mum disability and indicators of major chronic
conditions, the most djsabled patients, women,
those who were economically relatively inde-
pendent, those who lived with others, and those
who had been moderately active socially before
the present disabling illness.

In summary, we detected the times of appear-
ance of treatment effects during interval evalua-
tions; and the design of the study permitted us
to distinguish between interval treatment effects
which were unbiased and those which were
subject to possible observer or interaction bias.
The evaluations revealed that the presence of
observers did not bias nursing program effects,
while some true nursing program effects were
obscured by an interaction bias.

References

1. Chiang, Chin Long: Introduction to Stochastic

Process in Biostatistics. Published by John Wiley &
Sons, Inc., N. Y., N.Y, pp. 269-290, 1968.
CHAPTER XI.

What significance can we attach to the findings
of this study? To answer this question we must
relate our findings to information from other
sources, and to the parts of the health care
process which deal with home care and nursing
of the chronically ill. This chapter will attempt
to place the results in these perspectives.

Before interpreting the findings, we must
consider what general population our study
population may represent. The process of selec-
tion of this sample and the composition of the
sample have been described in Chapter 3. At this
point, we must remind ourselves that we are
considering what happened to a group of
elderly, predominantly female, chronically ill
patients who returned to the noninstitutional
community from a chronic disease rehabilitation
hospital. The hospital is in the category of
specialty geriatric and chronic disease hospitals
listed in the Master Facility Inventory of the
National Center for Health Statistics (1). At the
beginning of the study, 221 such hospitals were
listed in the Inventory. Abington House was one
of two which served most of Greater Cleveland’s
population. Patients in Abington House were
similar to those in four other chronic disease
hospitals (2) and were similar to patients in U.S.
geriatric and chronic disease hospitals in that
they were more disabled on the average than
residents in nursing and personal care homes (3).
Patients in our study differed from others in
these comparison groups in one respect. Al-
though people from all socioeconomic classes
were included, a larger proportion had moderate
to good financial resources.

The admission criteria in today’s chronic
disease rehabilitation hospitals are such that
patients who are admitted are generally disabled
and judged to have a potential for rehabilitation.
Since not all who could benefit are afforded this
rehabilitation opportunity, not all are repre-

INTERPRETATIONS

sented in the present study. Those admitted to
such hospitals, however, use a large amount of
the nation’s health service resources; and this
group is represented in the study.

If we are to interpret the findings correctly,
we must also consider certain kinds of bias to
which they may be subject. The completeness of
our data was excellent (Chapter 5), and the
randomization process produced experimental
groups which did not differ in respect to
important characteristics at the beginning of the
study (Chapter 12). Another possible source of
bias which the study permits us to examine is
the influence of introducing observers separately
or in parallel with the nursing program. We refer
to results which are related to these two factors
as ‘“‘observation effects” and ‘“‘interaction ef-
fects”, respectively. As stated earlier, the term
“interaction effect” applies to an effect meas-
ured by the study design, and should not be
confused with the term ‘social interaction”
which is used to describe an element of social
function. “Observation and interaction effects”
are described, next, as a basis for understanding
better the final section, ‘effects of the nursing
program,” in which we summarize and discuss
the effects of the service program.

Observation and Interaction Effects

We identified outcome during the two-year
study period that was related to the presence of
an observer by contrasting the 150 patients who
were assigned for quarterly observer visits with
the remaining 150 who were observed only at
the beginning and end of the study (Table 2.1).
These findings have a bearing on the general
problem of observational influence in studies of
human behavior and are summarized here for
the purpose of distinguishing observation effects
from the effects of the nursing program.

With regard to function, injury, and mortality,

83
Effects of Continued Care

we did not find observation effects in com-
parisons between the 150 patients who were
assigned for quarterly observer visits and the
remaining 150 who were observed only at

the beginning and end of the study. Among
subgroups which were homogeneous for disease,
disability, and mental function at the time of
entry into the study, observation effects were
distinctly less numerous than the nursing effects
and quite different from them in character and
direction (Figure XI-1). Compared to the nurs-
ing program, the process of observation had
little effect on physical or mental function.
Observation did, however, influence social func-
tion and admissions to hospitals, the principal
effects being ‘fewer decreases in social inter-
action” and “fewer patients admitted to hospi-

tals.”
Although we designed the study to permit us

to identify and measure observer effects, we did
not design it to explain the basis for such
effects. We consider it nevertheless appropriate
to present a simple explanation suggested by our
experiences in the study and by its findings,
which could lead to a better understanding of
the observation effects. The virtual absence of
observation effects on physical and mental
function probably reflected successful avoidance
of a professional service role by the observer, a
role which was strongly restricted by the train-
ing and supervision program for observers. How-
ever, the observer’s visit and her interest in
patient activities created an interactive situation,
which was an unavoidable stimulus for social
interactions. The observer who visited re-
peatedly and regularly may thus have taken on a
secondary role, namely, that of a nonmedical
friend, whose influence substituted social per-
spectives for medical or nursing perspectives
which then affected decision-making concerning
patients’ problems. Since the observation proc-
ess tended to maintain social interaction and to
decrease hospitalizations, this explanation ap-
pears to offer an appropriate perspective from
which to develop hypotheses about the observa-
tion effect.

The observer and care teams were Kept
separate throughout the study; yet, some sta-
tistically significant ‘interaction effects” ap-

84

peared. We identified these effects by contrast-
ing the two groups in which the nursing program
and observation process were parallel (N+O+ and
N-O- groups as defined in Table 2.1) with the
two in which the programs functioned independ-
ently (N+O- and N-O+ groups). The interaction
effect, therefore, was the study outcome which
resulted from the combined influences of the
nursing program and the observation process and
which was not attributable to either in the
absence of the other. Since treatment and
observation were separate geographically, this
effect is probably an indirect one as, for
example, in the instance of a patient who
discusses his problem with both a nurse and an
observer and then acts differently than he would
if he had talked to only one. Unless otherwise
stated, we discuss in this section those results
which were consistently associated with the
presence of both programs (N+O+ component of
the interaction effect).

Considering, first, the interaction and nonin-
teraction groups which included all patients
(150 in each group), we did not find differences
in physical function, mental function, injury,
and mortality. In terms of socioeconomic func-
tion and use of services, however, fewer in the
interaction group returned to employment
during the study than in the noninteraction
group (Figure XI-2). Those in the interaction
group also had more contacts with physicians
and had more home visits by physicians during
the last two weeks of the study.

Among subgroups which were homogeneous
for disease, disability, and mental function at
the time of entry into the study, the results of
interaction comparisons show that, whereas the
nursing program distinctly influenced physical
and mental function, the presence of an inter-
action between the nursing program and obser-
vation process did not have such effects. (The
first part of this section reported that observer
effects were similarly not demonstrated for
physical and mental function). Interaction ef-
fects were apparent, however, in the areas of
social function and use of services, namely,
fewer returns to employment, more physician
services, and less non-nutse attendant care (Fig-
ure XI-2). A possible interpretation is that the
Types of Patients’

Musculoskeletal Disease
without other chronic condition

Central Nervous System or
Cardiov lar-Ren

with other chronic condition

Severely Disabled
major assistance with self care

Not Severely Disabled
with cardiovascular-renal disease or cancer
without cardiovascular-renal disease or cancer

without dementia

Sex
male

 

Economic Status
independent

Household Composition
lives alone

Sociai Interaction
active

Interpretations

Types of Effects’

fewer hospitalizations

fewer decreases in social interaction
fewer decreases in social interaction
more decreases in social interaction; fewer hos-

pitalizations

fewer deteriorations in limb movement; fewer hos-
pitalizations

more decreases in social interaction; fewer hos-
pitalizations

fewer decreases in social interaction

fewer decreases in social interaction

fewer hospitalizations

fewer deteriorations in limb movement; fewer de-
creases in social interaction; fewer hospitalizations

Classes of patients are defined in Chapters 6, 8, and 14. Types of effects are defined in Chapters S and 14. Listed effects are those
which were most frequently encountered, that is, any type of observation effect which was statistically significant for at least five per
cent of the study’s Intake Classes (Chapter 14). Listed classes of patients are only those which showed these frequent effects.

Figure XI-1. Effects of the Observation Process

presence of both the nursing program and
observation process tended to reinforce the sick
role, as defined by Parsons (4). The fact that
patients in this situation were less likely to
return to work suggests that they felt “exemp-
tion from normal social role responsibilities,”
while the continued active contact with physi-
cians could have arisen from their socially
induced ‘obligation to seek technically com-
petent help.” This latter effect was in the same
direction as the effect of the nursing program,

and may well represent a common element in
the social and professional attitudes of ob-
servers and nurses, respectively. We consider,
also, that social stimulation was a contributory
factor to the interaction effects.

For a few subgroups, two findings were
associated with the absence of both the nursing
program and observation process (N-O- com-
ponent of the interaction effect) and could,
thus, not be interpreted as findings due to the
presence of interaction. These outcomes were:

85
Effects of Continued Care
Types of Patients’
All Patients

Musculoskeletal Disease
with other chronic conditions

Severely Disabled

major assistance with self care

Not Severely Disabled

without cardiovascular-renal disease or cancer

without dementia
no or minor assistance with self care

Age
65-74 years old

75 years old or older

Economic Status
independent

dependent

Social Interaction
active

inactive
Social Deprivation

not deprived

deprived

Types of Effects’
fewer returns to employment; more physician contact
last 2 weeks; more physician home visits last 2 weeks
more frequent physician home visits

less frequent non-nurse attendant care

fewer returns to employment; more increases in
economic dependence; more frequent physician
contact

more frequent physician contact

more frequent physician home visits

less frequent non-nurse attendant care

more frequent physician contact; more frequent
physician home visits

fewer returns to employment; more frequent physi-
cian contact

more frequent physician home visits; less frequent
non-nurse attendant care

fewer returns to employment; more frequent physi-
cian contact

more frequent physician home visits

more frequent physician contact; less frequent non-

nurse attendant care

fewer returns to employment; more frequent physi-
cian home visits

Classes of patients are defined in Chapters 6, 8, and 14. Types of effects are defined in Chapters 5 and 14. Listed effects are those
which were most frequently encountered, that is, any type of interaction effect which was statistically significant for at least five per
cent of the study’s Intake Classes (Chapter 14). Listed classes of patients are only those which showed these frequent effects.

Figure XI-2. Interaction Effects

86
more deteriorations in limb movement and in
observation and clear thinking. It is interesting
to speculate that these adverse effects may have
resulted from the absence of systematic and
objective decision-making which nurses or ob-
servers would introduce when evaluating patient
problems.

Effects of the Nursing Program

The objectives of medical and nursing services
are to maintain health, to prevent illness and its
consequences, to treat illness, and to restore
well-being. For patients who are chronically ill,
services differ in important ways from services
for those who have short-term diseases which are
only infrequently associated with persistent im-
pairment. We do not currently expect to achieve
primary prevention (prevention of onset) of
chronic disease. A more realistic goal is second-
ary prevention, that is, slowing of the disease
process after it has begun and prevention of
complications. In the treatment of chronic
disease, preventive, therapeutic, and restorative
services generally overlap; and continuing super-
vision is needed. The threats of chronic disease
are physical, psychological, and social. Corre-
spondingly, the services required are interdis-
ciplinary; and interdisciplinary decisions and
resources must be integrated through a cooper-
ative process of assessment, service, reassess-
ment, modification of service goals, follow-up,
and follow-through. It is within the framework
of these concepts that we discuss the effects of
the nursing program in this section.

In Chapters 7, 9, and 10, we described in
detail the changes in patient function and the
changes in use of services which were attribut-
able to the nursing program. In the present
chapter, such changes are summarized in quanti-
tative terms in Tables 11.1 through 11.4, which
supplement the antecedent material. We shall
discuss these results from three perspectives:
(1) the specificity of consequences with respect
to the type of treatment program and type of
patient, (2) the consequences of the treatment
program as an expression of its coordinating
activities, and (3) the consequences in terms of
social function. In this discussion, we interpret
results with the caution required by the method-

Interpretations

ology of the study, including appropriate
attention to the research strategy, availability of
research measures, selectivity of research meas-

ures, and limitations of statistical inference.
The nursing program had specific conse-

quences for patients with particular character-
istics, and the types of consequences were
related to the types of nursing program activ-
ities. A striking fact which is brought out by
contrasting Table 11.1 with 11.2 is that the less
disabled, less severely ill patients tended to show
physical and psychological benefits, while the
most disabled and severely ill received increased
professional services. Younger patients were
benefited in the same way as the less disabled, as
were also those who functioned adequately
mentally and those with uncomplicated muscu-
loskeletal disease. Older patients received more
services as did those with complicated central
nervous system or cardiovascular-renal disease.

We are reminded by these observations of
Parsons’ statement that medical practice is a
“mechanism in the social system for coping with
illnesses of its members” and that the role of
practitioners is ‘‘institutionalized about the tech-
nical content” of its activities or functions (5).
Parsons uses the term ‘‘functionally specific” in
referring to this concept, and findings in the
present study are consistent with the concept.
Both in the hospital where the study began and
in the nursing program, special skills and re-
sources were available for rehabilitation. We
would expect, under these circumstances, that
some of the disabled patients who were served
could benefit from such resources, while others
could not. We might also expect that specific
benefits from the available specific treatment
would be demonstrated most readily for those
who were relatively less disabled and, thus,
potentially more able to respond. Adaptation of
service to the needs of the patient and to his
capacity to respond is an important skill of the
health professional. In a study of home nursing
care, this was recognized by Johnson who
referred to adaptation of service as the ‘“‘differ-
ential opportunity open to the nurse” (6).

The findings of the study strongly confirm
the foregoing assumptions and have important
implications for several parts of the health care

87
Effects of Continued Care

Table 11.1 —Quantitative Estimates of Nursing Program Effects on
Function for Patients Classified by Disease, Disability, and Mental Status!

 

 

 

 

 

Characteristics Beneficial penafic] al Beneficial Decreased Increased
. effect in effect in .
on Admission walking . . . social house-
limb orientation and | . . .
to Study effect interaction | confinement
movement mental control
All Patients — ee — — —
Musculoskeletal Disease
with other chronic condition —_ — —- 31 —
without other chronic condition 26 17-34 35 — —
arthritis — 47 — — —
Central Nervous System or
Cardiovascular-Renal Disease
with other chronic condition —- — —_ — —
without other chronic condition — — — 60 —
Severely Disabled
major assistance with self care — - — 45 —-
confined to house — — — — -
unable to walk alone —- — — —_ -
restricted limb movement — — — — —
Not Severely Disabled
with cardiovascular-renal disease
or cancer — 14-27 — — -
with dementia — — — 45-57 —
without cardiovascular-renal disease
or cancer 16 16-17 — — —
without dementia — 20 34 — —
no or minor assistance with self care — 13-25 27 — —
not confined to house — 10-12 — — —-
walks alone — — 25 — —
unrestricted limb movement — 13 — — —

 

 

 

 

Ipatient characteristics and program cifects are those identified in Chapter 7.

Explanatory Note to Tables 11.1, 11.2, 11.3, and 11.4:

 

Each number in the following Tables represents the difference
between the per cent who showed a specified outcome among
those referred to the nursing program and the per cent who
showed that outcome among those not referred to the program.
In Table 11.1, for example, the figure of 47 per cent at the
intersection of the row labelled “arthritis” and the column
labelled “beneficial effect in limb movement” is obtained as
follows. In the group of patients with arthritis who were assigned
to the nursing program, 18 per cent showed deterioration in limb
movement by the end of the study, while in the group of
comparable patients not assigned to the nursing program, 65 per
cent showed such deterioration. The difference, 47 per cent, is

88

taken as an estimate of the beneficial effect attributable to the
nursing program. This figure means that, at the end of two years,
approximately half the patients with arthritis who were assigned
to home nursing care maintained limb movement which they
would have lost had they not been given this care. Most of the
results represent a difference between status at intake and at the
end of the study and, therefore, refer to two-year survivors. For
those outcome measures which were expressed as dates of
specific events (e.g., injury, admission to nursing home, admis-
sion to hospital), both survivors and non-survivors were included;
and the interpretation of such results considers whether or not
the events occurred before death or the end of the study,
whichever came first.

 
 

Interpretations

Table 11.2—Quantitative Estimates of Nursing Program Effects on Use of Services
for Patients Classified by Disease, Disability, and Mental Status!

 

Fewer
Admitted to
Nursing Homes

Characteristics
on Admission
to Study

More
Admitted to
Hospitals

Increased Use of Professional Services

 

At Time of Death

 

Physicians | Others

Hospital Care | Autopsy

 

All Patients en

Musculoskeletal Disease
with other chronic condition —-
without other chronic condition —-
arthritis -

Central Nervous System or
Cardiovascular-Renal Disease
with other chronic condition —
without other chronic condition -

Severely Disabled
major assistance with self care -
confined to house —
unable to walk alone —
restricted limb movement —-

Not Severely Disabled
with cardiovascular-renal disease
or cancer 20-21
with dementia —
with cardiovascular-renal disease
or cancer —
without dementia 20
no or minor assistance with self care 21
not confined to house -
walks alone 25
unrestricted limb movement —

 

 

12 — —~ ~ 41

28-48 - - - -

-_ 27-43 32-40 51 —

17-30 19-20 23 45

— 28 — — _

 

 

 

 

 

1p,tient characteristics and program effects are those identified in Chapter 7.

system. Specific types of patients who benefited
were identified; and terms of benefit were
identified as, for example, “maintain strength,
range of motion, and orientation and mental
control.” We, also, established quantitative esti-
mates of the beneficial effects, in general, 10 to
47 per cent for physical function and 24 to 35
per cent for mental function (Tables 11.1 and
11.3). By thus adding specificity to existing
concepts, our study contributes information
about the directions and limits of effectiveness

cof the nursing program. Study findings help

nurses identify the types of patients who were
benefited and the types who both were sup-
ported without apparent benefit and used more
services. Efficiencies are, thereby, suggested for
planning and allocating highly skilled and less
skilled nursing services. Nurses should delegate
many supportive service tasks to less highly
trained new health personnel, and this study
characterizes the likely categories of patients to
be assigned to skilled and less skilled assistance.

89

 
Effects of Continued Care

Table 11.3—Quantitative Estimates of Nursing Program Effects on Function
for Patients Classified by Age, Sex, and Social Characteristics!

 

Beneficial
walking
effect

Characteristics
on Admission
to Study

Beneficial
effect in
limb
movement

Beneficial
effect in
orientation and
mental control

Increased
house-
confinement

Decreased
social
interaction

 

Age
50-64 years old —

Sex

male _ _
female — _

Social Class

high _ _

low —_ —

Economic Status

independent —- —
dependent — —

Household Composition

lives with spouse — —
lives with other than spouse — —
lives alone — —

Social Interaction

active _ 36
inactive — —

Social Deprivation
not deprived —

 

 

 

20-24
65-74 years old — -
75 years old or older — —

15-19
deprived — _

25 —- —

a - 19

— — 32

- 22-29 -

 

 

 

 

patients characteristics and program effects are those identified in Chapter 9.

As a result, public health nurses could extend
their capacity to fulfill the community’s need
for professional manpower which is in short
supply. An implication of the findings for
education is that a clear recognition of the kinds
of professional goals which can be achieved with
different types of patients should help simplify
and rationalize professional education.

At the stage of chronic illness which most of
the patients in this study had reached before
being referred for care, goals of care were to
interrupt or slow the regressive processes which

90

had already begun. This situation was recognized
in the wording of the original hypotheses
(“maintain or increase function’). Our results
showed that maintenance of function or, more
precisely, avoidance of deterioration was, in
fact, the most consistent favorable effect, and
that even this result could be achieved only with
the younger and less disabled patients.

As previously stated, many patients with
chronic illness are so severely disabled that the
goal of treatment must shift from improving
or maintaining physical and mental health to

 
Interpretations

Table 11.4—Quantitative Estimates of Nursing Program Effects on Use of Service for

Patients Classified by Age, Sex, and Social Characteristics!

 

Characteristics
on Admission
to Study

Increased Use of Professional Services

 

More
Admitted to
Hospitals

Fewer
Admitted to
Nursing Homes

Physicians | Others

At Time of Death

 

Hospital Care | Autopsy

 

 

Age
50-64 years old
65-74 years old
75 years old or older

Sex
male
female

Social Class
high

low

Economic Status
independent
dependent

Household Composition
lives with spouse
lives with other than spouse
lives alone

Social Interaction
active
inactive

Social Deprivation
not deprived
deprived

23-24

 

 

20

33

19-20

30

22

28

24
70

28 46-49

 

 

 

 

 

patients characteristics and program effects are those identified in Chapter 9.

coping with their everyday needs and periodic
crises. It is not surprising that, for a group of
patients who are unable to perform independ-
ently such ordinary activities as bathing, dress-
ing, and using the toilet, the efforts of nurses
and others around them are directed mainly at
management rather than treatment. If the na-
tural course of events cannot be altered for such
patients, the public health nurse can at least
assure that they have access to other profes-
sional services when crises occur and when

regression proceeds. For the severely disabled
patient, then, the public health nurse is expected
to increase the patient’s access to hospitals,

doctors, and other health workers.
Posman and co-workers conducted a prospec-

tive, experimental study of public health nursing
services for patients over the age of 60 dis-
charged from a public chronic disease hospital.
In this group of patients, who differed from
those in our study mainly in social and eco-
nomic status, Posman identified a large disabled

91

 
Effects of Continued Care

subgroup for whom no favorable effects on
health, physical functioning, or employment
could be demonstrated but for whom the nurses
were instrumental in improving access to hospi-
tals and other professional services (7).

The effects of the nursing program on use of
services in our study are summarized in
Tables 11.2 and 11.4. Use of hospitals and
services of other health professionals was in-
creased in the presence of the nursing program,
especially for persons who were older, more
disabled, and more severely ill. Among these
people, quantitative estimates of increased use
of physicians and hospitals varied from 17 to 48
per cent, generally being 25 per cent or greater.
Autopsies were also more frequent among such
individuals, probably secondary to increased
hospital admissions. Admissions to nursing
homes, on the other hand, were less frequent
among those in the nursed group, particularly
those who were less disabled and who, as
indicated previously, avoided deterioration more
successfully when they were exposed to the
nursing program (20 to 25 per cent fewer
admissions in Tables 11.2 and 11.4). The per-
centage of days spent in nursing homes was
correspondingly less for the treated group. These
outcomes with respect to use of services were
consequences of the presence of the nursing
program, and they illustrate the selective appli-
cation of coordinating activities, confirming the
“functional specificity’ of the program which
we discussed earlier in this section.

The management of chronic illness is inter-
disciplinary, and the function of the public
health nurse as a coordinator of services has
been recognized in official statements as well as
in practice (8). One follow-up study of rehabili-
tation patients in New York City showed that
they and their families had great need for help in
finding and coordinating services, including care
by physicians and hospitalization (9). A com-
mon criticism of the health care system in the
United States is that it is badly fragmented (10).
The elderly, chronically ill individual at home is,
furthermore, particularly likely to be unaware of
or to lack access to persons and institutions who
might care for him. The public health nurse is in
a strategic position to help such patients make

92

contact with doctors, other health services, and
hospitals. Our study reveals the circumstances in
which interdisciplinary contacts of this kind
were established. Since extended professional
services were obtained primarily for patients
with the greatest limitations and those least likely
to demonstrate beneficial outcomes, it appeared
that the functions of professional services for
these patients were not sought primarily for
physical cure or amelioration. In this connec-
tion, we note that increased use of professional
services appeared six to twelve months after the
treatment program was introduced. The findings
suggest that such services were required in order
to manage the crises and regressive processes that
occurred, as well as to share the responsibility
for difficult decisions and for providing burden-
some care. For example, reports of the nurses’
activities showed a high proportion of psycho-
social care (Chapter 4) which consisted, in many
cases, of counselling the patient and family
about such major decisions as hospitalization or
admission to a nursing home. If we accept this
interpretation, it implies that services for this
type of patient should be organized with less of
the traditional emphasis on diagnosis, treatment,
and consultation and more on planning, guid-
ance, and interdisciplinary communication.
Viewed thus, it may quickly become apparent
that a balanced approach to service for many
chronically ill persons should include home-
makers, “sitters,” friendly visitors, and day care
centers, as well as doctors, nurses, and hospital
beds. Current emphases on doctors, nurses, and
hospital beds may, in fact, be out of balance
when measured against needs for social and
psychological supports. The need for “institu-
tional” care of the totally disabled will remain
with us; but this could be provided less expen-
sively and more humanely if, after adequate
evaluation and trial, certain wasteful efforts
were set aside and the provision of decent living
conditions substituted, with easy access to pro-
fessional services for crises.

In the present study, decreased social inter-
action and increased house-confinement were
outcomes which we attributed to the presence
of the nursing program. Decreased social inter-
action was experienced mainly by the oldest
patients, those receiving much assistance with
self care, those with cardiovascular-renal disease,
those with poor mental function, socially de-
prived patients, and economically dependent
patients (Tables 11.1 and 11.3). Since social
interaction is a complex phenomenon that has
its origins in an extensive combination of physi-
cal, psychological, social, and economic factors,
the direction of change in the interaction (in-
crease or decrease) does not indicate clearly
whether well-being of the patient increases or
decreases. Though increased social interaction
may, at first glance, appear to reflect increased
well-being, restricted interaction seems more
appropriate for many very disabled people.
Senile people, for example, require protective
geographic restriction; and chronically ill and
very old people may need protection to prevent
accidents, such as falls which cause hip fractures.
Social interaction for the infirm is often pre-
dominantly concerned with meeting their physi-
cal needs and with protection. Though this may
be interpreted as proper “disengagement” from
activities which are appropriate to younger
people (11), our societal behavior tends to em-
phasize the physical network of support more
clearly than the psychological and social net-
work. In this discussion of social interaction, we
have arrived at the same conclusion about social
and psychological needs of old and very disabled
people as we did in the discussion of increased
use of services for these people. The findings
also suggest that, among other risks, the profes-
sional process may reinforce inappropriate de-
pendency for certain patients.

The related finding among treated patients
was increased house-confinement for socially
active, economically independent men who lived
with a spouse. The composite of these character-
istics suggests a pattern of life in which the
occurrence of a major disabling disease (as did,
in fact, occur) would be a desolating event. In
addition, the same types of patients had in-
creased contacts with physicians and other
health professionals (Tables 11.3 and 11.4). A
discussion of the interpretation of these observa-
tions resembles the preceding discussion of
explanations about decreased social interaction
and increased use of professional services. Possi-

Interpretations

ble explanations in terms of treatment func-
tions, thus, include providing increased protec-
tion and reinforcing the patient’s sick role. In
addition, an active, economically independent
man who suffers a disabling illness requires a
different type of adjustment, for example, from
a housewife who may continue to be partially
involved in homemaking activities even after a
disabling illness. The opportunity to accept the
sick role may serve as the social process by
which this man redefines his place in society. We
are reminded, also, of Marden’s observation that
“impaired women possess more ‘insulation’
than men (12). Of interest is the observation
that increased use of professional services, de-
creased social interaction, and increased house-
confinement appeared concurrently six to
twelve months after the nursing program was
introduced. Our explanations of these outcomes
suggest that professional services may be asked
to substitute for social and psychological sup-
ports which society has not provided. We ques-
tion whether professional services alone are the
best way of meeting these needs and suggest, as
stated earlier, that less skilled assistance should
be assigned to certain categories of patients.
Continued care at home, the subject of this
book, is an important element of neglected
primary services for a segment of the population
which cannot speak effectively about its needs
and which uses a large amount of the nation’s

health service resources. For a field which needs
development, material in the book about the

expected results of home care adds to the base
of fact on which rational planning and action
should proceed. The information will, also, be
useful to practitioners and teachers. Use of
the experimental method in a manner which is,
thus far, unique in studies of “‘real world” health
service programs adds a large measure of confi-
dence to the results. In the remainder of the
book (Chapters 12, 13, and 14, and the Ap-
pendix), we describe the study methods in
expanded form.

References

1. National Center for Health Statistics: Health Re-
sources Statistics, Health Manpower and Health
Facilities, 1969. Public Health Service Publication

93
Effects of Continued Care

94

No. 1509, 1969 Edition,
237-242, May 1970.

Washington, D.C.,

. Goldman, F.: Patients in Chronic Disease Hospitals:

A Profile. A. J. P. H. 52: 646-655, April 1962.

. National Center for Health Statistics: Characteristics

of Residents in Institutions for the Aged and
Chronically Ill, United States, April-June, 1963.
Series 12, No. 2. Published by the U. S. Department
of Health, Education, and Welfare, Public Health
Service, Washington, D.C., pp. 12-15, September
1965.

. Parsons, T.: The Social System. Published by the

Free Press, Glencoe, Ill., pp. 436437, 1951.

. Parsons, T.: The Social System. Published by the

Free Press, Glen. se, Ill., pp. 432435, 1951.

. Johnson, W. L.: Content and Dynamics of Home

Visits of Public Health Nurses, Part II. Published by
the American Nurses’ Foundation, Inc., 10 Colum-
bus Circle, New York, N.Y. pp. 6-9 and 122-124,
1969.

. Posman, H., Kogan, L. S., Le Mat, A. F., and

Dahlin, B.: Continuity in Care for Impaired Older
Persons: Public Health Nursing in a Geriatric Re-
habilitation Maintenance Program. Report prepared

10.

11.

12.

for the Committee on Health by Harry Posman,
Department of Public Affairs, Community Service
Society of New York, 105 East 22nd St., New
York, N.Y., 126-148, 1964.

. American Nurses’ Association: Functions, Standards

and Qualifications for Practice of Public Health
Nursing. Published by the American Nurses’ Associ-
ation Public Health Nurses’ Section, 10 Columbus
Circle, New York, N.Y., pp. 3-16, 1962.

. Kelman, H. R., Muller, J. N., Lowenthal, M.: Post

Hospital Adaptation of a Chronically Ill and Dis-
abled Rehabilitation Population. J. Health Hum.
Behav. 5: 108-113, 1964.

White, K. L.: Organization and Delivery of Personal
Health Services: Public Policy Issues. Milbank Mem.
Fund Quart. 46: No. 1, Part 2, 225-256, January
1968.

Cummings, E. and Henry, W. E.: Growing Old: The
Process of Disengagement. Published by Basic
Books, Inc., New York, N.Y., pp. 47-51 and 244-
250, 1961.

Marden, P. G. and Burnight, R. G.: Social Conse-
quences of Physical Impairment in an Aging Popula-
tion. Gerontologist 9: 39-46, Spring, 1969.
 

CHAPTER XII.

This chapter is concerned with the statistical
aspects of the study and the drawing of valid
conclusions therefrom. The methods of statisti-
cal analysis and the strength and validity of the
conclusions depend greatly on the experimental
design and the rigidity with which we followed
this design. Accordingly, the experimental de-
sign described in Chapter 2 is re-examined in this
chapter from a more technical point of view. We
shall, also, examine the extent to which the
execution of the study actually conformed to
the experimental design, the methods of analy-
sis, and their relevance to the purpose of the
study and to the experimental design.

The Experimental Design

The successes achieved by the science of Physics
through the experimental method have largely
resulted from the physicist’s ability to identify
and manipulate, at his pleasure, important vari-
able factors in his experimental studies. The
physicist often has the advantage of working
with material that is homogeneous. His ability to
express results quantitatively and in the form of
equations allows him to predict results accu-
rately under a variety of circumstances, and to
build up gradually and increasingly inclusive
theoretical framework.

The biomedical sciences must deal with the
huge amount of variation inherent in biological
organisms. Since no two people are exactly
alike, a treatment that is successful for one
patient may not be successful for another
patient with the same illness. The multiplicity of
factors affecting individuals in various degrees,
and at various times, all too frequently makes it
exceedingly difficult to identify important indi-

PART Ii

methods

STATISTICAL METHODS

vidual factors. For studies involving people,
social customs, laws, and ethical considerations
often prevent the experimental manipulations of
factors suspected to be of importance.

In experiments involving people, then, it is
usually impossible to identify and control ade-
quately all the important extraneous factors
capable of affecting the experimental results.
Fortunately, when the experimental method is
to be applied, there is recourse to the process of
randomized allocation of subjects. By this
random process, the various factors, important
and unimportant, known and unknown, are each
expected to occur in equal proportions in the
experimental groups. Their influences on out-

come will, on the average, be the same in each
group; and a scientific assessment of the rela-

tionships between treatments and outcomes is
possible.

We assigned each of the 300 patients enter-
ing the present study, at the time of his entry
and strictly on a random basis, to one of four
groups which were referred to as “primary
groups.” These consisted of the four possible
combinations of the two experimental variables,
namely, the nursing program and the observa-
tion process. We adopted the symbols shown on
the following page for the primary groups.

The 150 patients in the N+O+ and N+O-
primary groups were scheduled to be visited
regularly at home by a VNA nurse, while those
in the remaining two primary groups were not.
Every patient in the study was scheduled to be
interviewed at “intake” when he entered the
study, and at ‘“‘termination” when he left the
study two years later. In addition, the 150
patients in the N+O+ and N-O+ primary groups

95
Effects of Continued Care

Combination of
experimental variables

 

 

Primary Referred Quarterly Number in
group to nursing observation primary
symbol program process group
N-O- absent absent 75
N+O+ present present 75
N+O- present absent 75
N-O+ absent present 75

were scheduled to be interviewed at home every
three months for a period of two years and were
scheduled to have a post-discharge interview one
week after entry into the study. The 150
patients in the other two primary groups were
not scheduled for such interval evaluations. With
few exceptions, each of the 150 patients in the
N+O+ and N-O+ primary groups had the same
interviewer throughout the study, but had a
randomly assigned different interviewer for the
terminal interview. Each of these 150 patients
was scheduled to be interviewed twice at the
end of the study, once by the randomly assigned
new observer and a week later by the regularly
assigned observer, for reliability and interval
comparison purposes. Since patients shared com-
mon reference points in time (entry, quarterly
observations, termination), their dates of entry
could be transformed to a common starting or
zero point; and all subsequent events on the
time scale could be referred to the zero point.
The patients thus comprised a cohort, and those
in prospectively defined subgroups comprised
subsets of the cohort.

The process of random assignments to the
primary groups guaranteed that each patient had
an equal chance of being assigned to any of the
groups. It also guaranteed that extraneous fac-
tors (large or small, important or not, known or
unknown) would, on the average, be equally
distributed among the four primary groups. The
practical consequence of the foregoing theoretic
statements is that patients in the four groups
were expected to be as similar as possible. We
expected, for example, that patients with his-
tories of serious heart disease would be equally
distributed among the four groups, as would

96

men or those who preferred radishes to cu-
cumbers.

Random assignments were accomplished with
the aid of a table of random numbers. For each
of 15 successive blocks of 20 patients, the
randomization process equalized the numbers of
patients in the primary groups. We designed this
sequential blocking to prevent extended series of
patients being assigned to the same study groups
over the 22-month intake period. Sequential
blocking, also, permitted us to examine findings
in successive groups over time in order to detect
differences in findings which might be associated
with such external changes as new treatments,
environmental changes, and new legislation such
as Medicare. (Comparisor.s of findings for the
first patients entering the study against findings
for the last did not reveal marked differences.)

As we described in Chapter 2, the four

equal-sized primary groups were pooled in vari-
ous combinations, so that we could distinguish
several types of effects. The following combined
groups, called ‘“‘experimental groups’, were de-
fined as shown on the following page.
Since the primary groups were formed at ran-
dom, it follows that the experimental groups
were also constructed in a random manner. Each
patient, thus, had an equal chance of being
assigned to any of the experimental groups, and
extraneous factors would, on the average, be
equally distributed among the experimental
groups, e.g., age, sex, state of health, etc.

Comparisons between experimental groups
were used to determine various effects in the

study, and we discuss such comparisons and
effects in the next section. The following effects

were defined as shown on the following page.
The reader will note that we adopted the follow-
ing format with respect to the use of symbols in
this chapter. Primary groups are designated by
both an N and an O, each followed by an
algebraic sign. Symbols for experimental groups
include a single algebraic sign which is not
enclosed by parentheses, while symbols for
effects are either without algebraic signs or with
such signs enclosed by parentheses.

Experimental Effects

The purpose of the study was to determine the
Statistical Methods

 

 

Experimental Experimental
group group symbol
referred to nursing program N+
not referred to nursing program N-
quarterly observation process O+
no quarterly observation process O-
“plus” interaction NO+
“minus” interaction NO-

Number in
experimental group

Combination of
primary groups

N+O+ with N+O- 150
N-O+ with N-O- 150
N+O+ with N-O+ 150
N+O- with N-O- 150
N+O+ with N-O- 150
N+O- with N-O+ 150

 

 

 

Experimental effect

Experimental effect symbol

Experimental groups compared

 

 

nursing program

observation process

interaction

nursing program in presence of observation process

nursing program in absence of observation process

N N+ against N-

Oo O+ against O—

NO NO+ against NO-
N(O+) N+O+ against N-O+

N(O-) N+O- against N-O-

 

consequences of the nursing program with re-
gard to various outcomes, at various times, and
for various people. If there were no interaction
between the nursing program and observation
process, the comparison of N+ against N- would
yield a valid measure of the nursing program
effect, since the presence or absence of system-
atically repeated observations was balanced
equally between the N+ and N- experimental
groups. Similarly, the comparison of O+ against
O- would yield a valid measure of the observer
effect, since referral or non-referral to the
nursing program was balanced equally between
the experimental groups O+ and O-. If an
interaction between the nursing program and
observation process did occur, the consequences
of the nursing program would not necessarily be
reflected in a comparison between the N+ and
N- experimental groups, since interaction be-
tween the nursing program and observation
process was a potential source of bias in one-
fourth of the sample (primary group N+O+) and
was not equally balanced between the experi-
mental groups.

Stated in another way, we must know
whether the nursing program and observer ef-
fects were independent of each other if we
would make interpretations about the conse-
quences of the nursing program from compari-
sons between those assigned and not assigned to
the program. Though the sole function of
observers was to gather data and though every
effort was made to ensure that nurses and
observers operated independently of one
another (separate housing, secretarial staff,
administrative staff, special liaison agents, etc.),
certain interacting effects could have occurred.
As stated earlier, a patient seen regularly by a
nurse and not interviewed regularly could have
behaved differently from a similar patient who
was seen regularly by a nurse and also inter-
viewed periodically.

When an interaction effect was present, the
interaction factor in the N+O+ primary group
invalidated the N+ against N- comparison as a
measure of the nursing program effect. A less
biased comparison was needed. The unbiased
comparison between the N+O- and N-O-

97

 
Effects of Continued Care

primary groups served this need, and such data
were available at the terminal point of the study;
however, this comparison could not be used for
quarterly interval data whose source, by defini-
tion, was the observers. Use of the terminal
unbiased comparison entailed a reduction in
total sample size from 300 to 150, a reduction
which was rarely required in the present study.
Significant effects of the nursing program based
on the unbiased comparisons are presented in
Chapter 11 and in the Appendix for 2 year
results, and in Chapter 10 for interval results.

Intake Classes and Effectiveness
of Randomization

Upon entering the study and before being
randomly assigned to study groups, patients
were systematically described by the observers.
Classes of patients with similar characteristics
(e.g., men, women, patients 75 years old or
older) were defined and were termed “intake
classes.” We used the intake classes to test the
effectiveness of the randomization process and
to compare experimental findings between simi-
lar subgroups of patients. Another set of classes
was established for use in identifying changes
due to the treatment program, i.e., “outcome
classes,” which we shall discuss later in this
chapter.

A total of 71 intake classes were established
prospectively as relevant to the study. We
discussed these physical, psychological, and so-
cial classes and their conceptual origins in
Chapters 6 and 8; and detailed definitions are
found in Chapter 14 (Intake Classes #1-#71). At
the end of the study, 19 other classes were
formulated in order to check the randomization
process and study findings. Such checks vali-
dated the process and findings and, unless
otherwise stated, results reported in this book
concern only the 71 intake classes which were
defined prospectively.

In order to determine the extent to which
small changes in intake class definitions affected
the study results, we designed certain intake
classes to be closely related to others. Overlap
between intake classes was evaluated with the
aid of correlation coefficients which were cal-
culated as follows:

98

r= [Z-DNA UZ)? [20-9)2])

Where, x = 1 or 0 when the patient is in or not
in the first class, respectively,
y =1 or 0 when the patient is in or not
in the second class, respectively,
x = the mean of the x’s, and
»y = the mean of the y’s.

The correlation coefficient, r, ranges between - 1
and +1. If r=-1, the two intake classes are
mutually exclusive (no overlap), and patients in
one of these classes are not in the other. If r = 0,
then the two classes are uncorrelated, and
patients in one of these intake classes are
randomly in or not in the other class. If r = +1,
the two classes are identical (complete overlap),
and patients in one of these intake classes are
also in the other. In general, the greater the
correlation coefficient, the greater the overlap.

We calculated the correlation coefficients for
all possible pairs of the 71 intake classes and
considered a correlation coefficient of +0.75 to
indicate major overlap. By this definition, we
identified thirteen sets of intake classes where
each intake class in a given set was highly
correlated with at least one other intake class in
that set. The thirteen highly overlappings sets
are listed on the following page, and brief
descriptions are included in order to clarify the
nature of designed overlap.

As stated previously, we expected the ran-
domization process to produce similar primary
groups and similar experimental groups. We
expected, thus, that proportions of patients in
the individual intake classes or in combinations
of intake classes would be the same for the four
primary groups (N-O-, N+O+, N+O-, N-O+)
and for the three pairs of experimental groups
(N+ and N-, Ot and O-, NO+ and NO-). The
extent to which the randomization process was
effective was checked by chi-square tests applied
to contingency tables in which the distributions
of intake characteristics were compared between
study groups. For example, the distribution of
men and women among study groups is pre-
sented in Tables 12.1 and 12.2. Chi-square tests
applied to these tables revealed no significant
difference in the distribution of men and women
among the primary groups or between the
Intake classes with
designed overlap
(by Intake Class #)

Statistical Methods

Description of set
(see Chapter 14 for detailed definitions)

 

3,5,7,9,11,25,27

4,6,8,10,12

13,17

15,21

18,22

24,26

28, 30, 32,41

29,31,33,42

34, 36,38

35,37

53,68

59, 65

61, 67

 

same degree of mild to moderate disability:
different combinations of cancer and indicators of cardiovascular-renal disease

same degree of mild to moderate disability and no cancer:
without cardiovascular-renal disease by various combinations of indicators of such disease

same degree of mild to moderate disability:
different levels of mental function

same degree of mild to moderately severe disability:
different levels of mental function

same degree of dementia:
different degrees of severe disability

same type of cardiovascular-renal disease and no cancer:
different combinations of disability and dementia

same type of comorbidity:
different combinations of principal diagnoses of the central nervous system and cardiovas-
cular-renal system

same absence of comorbidity:
different combinations of principal diagnoses of the central nervous system and cardiovas-
cular-renal system

same type of comorbidity:
different combinations of principal diagnoses of musculoskeletal

same absence of comorbidity:
different combinations of principal diagnoses of lower extremity fracture

similar type of role deprivation:
different in terms of loss of spouse

same type of contacts with spouse, relatives, and friends:
different degrees of other social interaction

same absence of daily contacts with people:
different degrees of other social interaction

experimental groups (at statistical probability
levels of 0.05). As another type of example, we
present in Figure XII-1 the distribution of ages
among those assigned and not assigned to the
nursing program. The observed age distributions
were not significantly different.

Actual and expected percentages of statisti-
cally significant chi-square tests for intake class

characteristics are compared in Table 12.3. The
similarity between actual and observed percent-
ages demonstrates that the randomization proc-
ess was very effective in distributing the 71
prospectively defined intake characteristics and,
theoretically, other extraneous factors equally
among the study groups. Among the few differ-
ences which we identified, the 150 patients

99
Effects of Continued Care

Table 12.1 —Comparison of Numbers of Men and
Women Among Primary Study Groups!

 

 

 

Primary groups Men Women Totals
N-O- 22 53 75
N+O+ 30 45 75
N+O- 22 53 75
N-O+ 25 50 75
Totals 99 201 300

 

 

 

 

 

 

IMen and women are Intake Classes #43 and #44, respectively,
as listed in Chapter 14. Symbols are defined in the text (Chap-
ter 12).

Table 12.2—Comparison of Numbers of Men and
Women in Experimental Study Groups!

 

 

Experimental groups Men Women Totals
N+ 52 98 150
N- 47 103 150
O+ 55 95 150
O- 44 106 150
NO+ 52 98 150
NO- 47 103 150

 

 

 

 

 

 

IMen and women are Intake Classes #43 and #44, respectively,
as listed in Chapter 14. Symbols are defined in the text
(Chapter 12).

assigned to the nursing program had principal
diagnoses of cardiovascular-renal disease with
indicators of coexisting major chronic condi-
tions more often than those not assigned to the
program. In the area of social characteristics,
those assigned for nursing services were more
likely to be living with a spouse or child.
Patients who were observed regularly did not
differ from those who were not observed regu-
larly in respect to any of the 70 intake charac-
teristics. In the search for differences between
patients classified by interaction between the
nursing program and observation process, the
interaction group had fewer patients with hyper-
tension and fewer who lived alone. Whereas
nonrandomized studies are subject to the criti-
cism that the outcome may be a result of

100

HD
o
1

Referred to visiting nurse (N+)

Ol
0
T

 

ow
Oo
T

 

nN
[$4]
T

 

nN
Oo
T

 

 

 

 

 

o

 

 

 

NUMBER IN 5-YEAR INTERVAL
a
T

 

 

 

 

 

 

oOo Oo
T

 

100

[9]
Oo

60 70 80 90
AGE IN YEARS

H
o
1

Not referred to visiting nurse (N-)

Ww
wn
T

 

ow
o
r

 

 

nN
on
T

 

 

 

S
I

 

 

NUMBER IN 5-YEAR INTERVAL
n
Oo

 

 

 

 

 

 

 

 

|

 

50 60 70 80 90 100

AGE IN YEARS

Figure XII-1. Age Distribution on Admission into Study

differences in intake characteristics, our study is
not subject to this criticism.

Validity of the Experimental Groups

The fact that patients were assigned to the four
primary groups (N-O-, N+O+, N+O-,N-O+) by
the randomization process did not necessarily
mean that they would participate fully as
assigned. A patient assigned to the N-O+ group,
for example, was not necessarily deprived of the
nursing program, while one assigned to the
N+O+ group might not participate fully in the
nursing program and observation process. It was,
thus, necessary to examine the extent to which
those patients assigned to the N+ experimental
group actually participated in the treatment
program and, equally important, the extent to
Table 12.3—Expected and Actual Percentages of Signifi-
cant Chi-Square Tests for Intake Class Distributions

 

 

Among Study Groups!
Percentages
“Expected” 5 10 20 30
“Actual” for
Primary Groups 0 7 13 21

“Actual” for
Experimental Groups

N+ against N- 4 10 17 30
O+ against O- 0 4 14 29
NO+ against NO- 3 13 23 34

 

 

 

 

 

 

 

LExpected percentages are the proportions of statistically
significant results which correspond in theory to individual
chi-square values. Actual percentages are the proportions of
significant results which were observed at these chi-square
values. Symbols are defined in the text (Chapter 12).

which those assigned to the N- group did not. It
was, similarly, necessary to examine the extent
of participation in the observation process for
the O+ and O- experimental groups.

A detailed description of patient participation
in the nursing program was summarized in
Tables 4.1, 4.2, and 4.3 of Chapter 4, where we
concluded that (1) the number of people in the
nursed group who received nurse visits was much
greater than the number in the non-nursed
group; (2) of those receiving visits, the people in
the nursed group received more visits per person
than those in the non-nursed group; and (3) the
nursing care follow-up was more sustained in the
nursed group than in the non-nursed group. The
large difference in visits was also demonstrated
in Figures IV-1 through 1V-6. Finally, in a
separate statistical analysis, we demonstrated
that, throughout the study, every intake class
for the N+ group received significantly more
visits than the corresponding intake class for the
N- group.

The 300 patients in the study had access to
private nurses, companions, relatives, and other
persons giving care, as well as to the VNA. Data
on non-VNA assistance was obtained in inter-
views, and we found no significant differences
between nursed and non-nursed experimental

Statistical Methods

groups or subgroups (intake classes) with regard
to non-VNA assistance. Such additional care did
not, therefore, affect the validity of the experi-
mental groups.

In view of the above discussion, there are no
qualms about the validity of the N+ and N-
experimental groups. The N+ and N- groups can
be correctly said to “receive” and “not receive”
the treatment program, respectively.

In Chapter 5, we examined in detail the
extent of participation in the observation proc-
ess for the Ot and O- experimental groups.
Tables 5.1 and 5.2 in that chapter demonstrate
the marked success with which the observation
schedule was accomplished for patients in the
O+ group (assigned to quarterly observations)
and O- group (assigned only to intake and
terminal observations.) Although certain inter-
view forms had greater completion rates than
others, successful completion of the type
described in Chapter 5 was the rule. In no case
did we completely lose a patient to follow-up.
The few patients who moved out of the state
were contacted through close relatives or
friends, and their local physicians were con-
tacted to assist with required evaluations. More
than 93 per cent of the evaluations were
achieved within nine days after the dates they
were due.

We designed the randomization process to
insure that each interviewer was assigned the
same number of patients within each primary
group and at random. Individual observer effects
were, thus, distributed equally among the
primary and experimental groups, and were
more surely related to the observation process
than to the individual observers. The only
deviation was for reasons of illness of observers
or because of vacations. Illness or vacations
required substitute observers, none of whom was
a regularly scheduled observer. Substitutions
comprised 4 per cent of intake evaluations and
less thereafter. The disparity at intake was larger,
since the contents of the observer-assignment
envelopes were unknown at that time, and
planning with respect to vacations could not be
foreseen nor effected.

In summary, the above discussion confirms
the validity of the O+ and O- experimental

101
Effects of Continued Care

groups. We can reasonably interpret differences
between these groups as related to the observa-
tion process.

Outcome Classes

Classes were defined to evaluate the conse-
quences of the nursing program. These classes
were entitled “outcome classes” and were ex-
pressed in terms of function (physical, psycho-
logical, and social), injury, death, and use of
services. We present detailed definitions of the
outcome classes in Chapter 14. Whenever possi-
ble, outcome classes were defined to measure a
change in the patient from the time of entry
into the study. For example, Outcome Class #3
at six months included patients who deteri-
orated in activities of daily living during the six
months period after entering the study. This
technique of treating each patient as his own
control was conceptually equivalent to in-
creasing the homogeneity of intake classes, as
well as of primary and experimental groups.

Though we dichotomized the outcome classes
(e.g., dead or alive, deteriorated or not, admitted
to a nursing home or not), there were occasions
when membership in an outcome class could not
be determined due to missing or incomplete
information. In other instances, patients who
died were excluded by definition for conceptual
reasons as in the case of Outcome Class #4,
“patients who deteriorated in walking.” In yet
other instances, information was unknown
because the study was not designed to collect it,
as in the case of quarterly interval information
for non-observed patients or in the case of
information about social interaction which was
designed to be collected semiannually and not
quarterly. As discussed later in this chapter and
in Chapter 5, the completeness of the data was
excellent; and we made careful interpretations
which took into account the various types of
unknowns. With respect to death, the design of
the study was to analyze the occurrence of
death as one type of outcome and to exclude
those who died in analyses of other outcomes,
rather than to diffuse the interpretations about
such outcomes by including deaths. Users of the
data can, thus, clearly understand the bases for
our interpretations.

102

In order to reduce the number of unknowns
for the various outcome classes at the terminal
point of the study, 24 month data obtained by
observers assigned to quarterly evaluations were
substituted in the occasional instances where 24
month data could not be obtained by ran-
domly assigned new observers. We compared the
two types of data, before pooling, to see how
closely they agreed (Table 12.4). Agreement was
good and served as an indication of the repro-
ducibility of the observation process, recog-
nizing also that the following factors con-
tributed to the variability of such data. Though
both sets of data were obtained at 24 months,
the pairs of evaluations did not take place at
exactly the same time. The pairs of evaluations
were, furthermore, conducted by different inter-
viewers; and patient recall was not constant.

Statistical tests were used to check the proc-
ess of random assignment of patients to the
primary and experimental groups, by which
process one expects to balance out such ex-
traneous factors as age, sex, and degree of illness
between the study groups. Other uses of statisti-
cal tests were to check the validity of the study
design, to test associations which might have
been related only indirectly to the study design,
and to discover the nursing program effects on
outcomes. For these purposes, we constructed
contingency tables as comparisons between the

Table 12.4—Agreement Between Two Sets of 24
Month Data Obtained by Separate Interviewers

 

 

Outcome class! Per cent agreement
#3 94
#4 93
#5 93
#12 98
#13 100
#14 91
#15 87
#16 88
#17 99
#23 95
#24 87
#25 87

 

 

 

1outcome Classes are defined in Chapter 14.

 
various study groups. The test generally used to
determine whether there was a statistically
significant difference was the chi-square test. No
test of significance was done on a 2 bv 2 table if
any row or column totalled less than 4, or if n
was less than 13. We applied Fisher’s exact
method for calculating probability if n was less
than 20, or if n was less than 40 and the
expected value for any of the four cells of the
table was less than 5. We used the chi-square test
in all other instances. All tests were two-tailed.
In an experimental design such as the one used
here, statistical analyses do not determine with
complete certainty the presence or absence of an
experimental effect on any particular outcome.
The degree of uncertainty may, however, be
calculated and expressed precisely in terms of
probabilities. In the present study, we consid-
ered a result as statistically significant if the
relevant statistical test was significant at the 5
per cent level (P<<0.05). This means that, in the
absence of a nursing effect on a particular
outcome, we would incorrectly state that there
was a nursing effect 5 times out of 100 in
repeated performances of the experiment under
similar circumstances. Recognizing that, in
examining a very large number of tests, some
differences are likely to occur by chance, we
assessed and described the consistency of differ-
ences. We also compared the most frequently
encountered outcomes (any outcome which was
significant for at least five percent of the study’s
Intake Classes) with less frequent outcomes, as
reported in Chapter 7.

When the chi-square value for a given con-
tingency table is large enough to be statistically
significant (P<0.05), one of several interpreta-
tions is usually given: (1) the P-value is signifi-
cant because of a causal relationship between
the factors compared in the table; (2) the
P-value is significant because of an association
(not necessarily causal) between the factors
compared in the table; or (3) the P-value is
significant for reasons due solely to chance. The
appropriate interpretation or valid conclusion
that can be inferred from a P-value depends
most strongly on the design and execution of
the study. Several different ways of obtaining
the data were used in the present study, and it is

Statistical Methods

crucial to the interpretation of the study’s
results to understand them. Brief examples are
in order; and for this purpose, two models of
tables will be considered.
Case 1: The model table is
Admitted to a nursing home
during the study
(Outcome Class #18)

Primary Group Yes No
N+O- a b
N-O- c d

In this case, extraneous factors which influence
entry into a nursing home are expected to be
distributed equally between the primary groups
because of the randomization process. The
observer effect is absent; therefore, any unusual
distribution of the numbers (a, b, ¢, d) in the
table may be attributed to the effect of the
nursing program. A significant chi-square value is
validly interpreted here as evidence of a causal
relationship between the nursing program and
subsequent nursing home admission, and the
direction of the causal relationship may either
be negative or positive (fewer or more admis-
sions, respectively). By definition, a significant
chi-square value will occur 5 per cent of the
time, by chance, in the absence of a nursing
program effect on outcome. On the other hand,
a nonsignificant chi-square value may occur even
when there is a causal relationship. This may be
due to chance, to too small a sample size, to
relative insensitivity of the outcome measures,
to a relatively weak causal relationship, or to
some combination of these. The probability of
realizing a nonsignificant chi-square value in the
presence of an actual causal relationship cannot
be calculated because of the lack of precise
knowledge of the extent of the causal relation-
ship. Tables similar to the model table for Case 1
are interpreted in the same manner.
Case 2: The model table is
Admitted to a nursing home

(Outcome Class #18)
Experimental subgroup

(Intake Class #43) Yes No
N+ males a b
N- males c d

103
Effects of Continued Care

This case differs from the preceding one in
two important respects. First, only males are
considered, and patients were not randomly
assigned to experimental groups according to
sex. Second, males of all four primary groups are
considered, and the presence of observers in two
of the groups could lead to observer or interac-
tion biases. As stated earlier in this chapter, the
process of random assignment guarantees that
extraneous factors such as being a male are, on
the average, equally distributed between experi-
mental groups. Within the limits of statistical
inference, therefore, interpretations of causal
relationships are valid. With respect to possible
observer or interaction biases, the present study
was designed to identify such biases. When
found, we took such biases into account in order
to make valid interpretations.

We scheduled outcome evaluations at eleven
times during the two years of the study as
described in detail in Chapter 2. The N, O, and
NO effects were calculated for each of the 1988
combinations of the 71 intake and 28 outcome
classes at the end of the study. As we explained
in Chapter 10, dates of death, injury, admission
to nursing homes, and admission to hospitals
were identified for all 300 patients; thus, inter-
val treatment effects during the two years could
be determined exactly with respect to these
outcomes, and without bias by the observation
process. For other outcomes, the only interval
effect that could be determined was a nursing
effect in the presence of observers (N+O+
against N-O+). This effect, it will be recalled,
could have been biased by the observation
process; and we discuss its analysis and interpre-
tation in detail in other parts of this chapter and
in Chapter 10. In every case where an inter-
action effect could be determined and was
found to be statistically significant, the unbiased
nursing program effect (N+O- against N-O-)
was calculated and substituted for the N effect.
In every case where interaction effects could not
be determined exactly, nursing program effects
were examined in the light of the interaction
effects which were identified at the end of the
study; and limited interpretations about un-
biased nursing program effects were made on a
defined and rational basis.

104

As an extensive example of analysis and
interpretation, let us consider experimental ef-
fects in terms of admission to a hospital (out-
come Class # 19) during the first year of the
study for patients who had not recently retired,
ceased homemaking, or been widowed (Intake
Class # 69). The data and results of statistical
analyses are presented in Table 12.5. Chi-square
test results reveal statistically significant N and
NO effects at the 5 per cent probability level
and no significant O effect. The direction of the
N effect is positive (more patients admitted to
hospitals in the presence of the nursing pro-
gram), while the direction of the NO effect is
negative (fewer patients admitted in the pres-
ence of interaction). If there were no interaction
between the nursing program and observation
process, the comparision of N+ against N-
would yield an unbiased measure of the nursing
program effect. Since an interaction effect is
present, the nursing program effect represented
by the N+ against N- comparison is biased by
the presence of the observation process; and the
best unbiased comparison for the nursing pro-
gram effect is N+O- against N-O-. In Table 12.5,
this unbiased effect is 33 per cent and is statis-
tically significant at the 5 per cent probability
level. Its direction is positive, that is, more
patients were admitted in the presence of the
nursing program than in its absence. We conclude
that there was a causal relationship between the
presence of the nursing program and admission
to hospitals during the first year of the study.

We can define the effects in the preceding
example precisely in terms of percentage differ-
ences between the comparison groups for par-
ticular outcomes. The statistically significant
effects are 19 per cent for the N effect, - 16 per
cent for the interaction effect, and 33 per cent
for the N(O-) effect. As discussed previously,
the N(O-) effect (33 per cent) is the best
unbiased measure of the nursing program effect
in this example. The effects in terms of percent-
age differences are quantitative estimates of the
study’s consequences, representing proportional
changes which would be expected for surviving
patients of this type as a result of the service
program which was applied. Such quantitative
estimates were summarized in Chapter 11.
Statistical Methods

Table 12.5—-Experimental Effects on Hospitalization During the First Year of the Study
in Socially Nondeprived Patients!

 

 

 

 

 

 

 

 

Hospitalized
Study Experimental Chi-square
Group Yes No effect value
(number of persons)

N 0 N=19% 6.672
0 . a 081
No. po o NO =-16% 4.882
Nor a > N(O+)= 4% 0.12
No i” - N(O-) = 33% 10.792

 

Socially nondeprived patients are those who had not recently retired, ceased homemaking, or been widowed (Intake Class #69 in
Chapter 14). Hospitalized is the Outcome Class #19 in Chapter 14. Symbols for the study groups and experimental effects are defined

in Chapter 12.

Value is statistically significant at the 0.01 level.

105

 
CHAPTER XIII.

EVALUATION SCHEDULE, STUDY

FORMS AND SUPPLEMENTARY DEFINITIONS

This chapter includes: the chronological sched-
ule of evaluations for the study; all forms used
to accumulate information; and supplementary
definitions to the material presented in Chap-
ter 5, Interviewing and Measures. The chronolog-
ical schedule appears as Figure XIII-1, which
lists the title of each form and the times when
each was completed. The study forms which
appear next are worded in enough detail to
minimize the need for additional definitions;
however, observers needed access to additional
written material about definitions and pro-
cedures. Such additional information was made
available to observers in the form of a guide. The

106

guide included the particular interview instruc-
tions for this study, the content of the observa-
tions, and the published sources of the measures
described in Chapter 5. The guide, also, included
the material in the present chapter. Definitions
and procedures that appear in published litera-
ture are not reproduced in full detail, since this
would be unnecessary duplication. For those
measures, however, that are not adequately
described in published literature, definitions and
descriptions of procedures are presented as the
last section of this chapter, “Supplementary
Definitions”.
 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Post
Form! Information Obtained Intake | Dis- 3 6 ° 12 15 18 21 fermi 24
Mos. | Mos. | Mos. | Mos. | Mos. | Mos. | Mos. nal Mos.
charge

A-1 Dates Evaluation Due (Interview Schedule) X X X X X X X X X X X
A-2 Admission Socioeconomic Evaluation X
A-3 Social Interaction X X X X

TA-3 Social Interaction X X
A4 Social and Economic Function® X X X X X X X X X

TA-4 Social and Economic Function X X
A-5 Activities of Daily Living® X X X X X X X X X X X
A-6 Medical Care X X X X X X X X X

TA-6 Medical Care X X
A-7 Permanent Move Completed at Time of Permanent Move
A-10 Physical Examination, Range of Movement & Strength X

TA-10 Physical Examination, Range of Movement & Strength X
A-11 Laboratory Examinations & EKG Findings X
A-12 Principal Diagnosis Leading to Admission X
A-20 Q-Sort Psychosocial Test (Highland View) X X
A-21 Coloured Progressive Matrices Test (Raven) X X
A-22 Orientation and Mental Control Test X X
A-18 Data from Death Certificate Completed at Time of Death

TA-4,
6, 10  Socio-Medical History of Expired Patient Completed at Time of Death

Greater Cleveland Confidential Referral Form X | |

AR-1 V.N.A. Activity Record—E xercises Completed at Time of Every Nurse Visit

AR-2  V.N.A. Activity Record—Prescribed Treatment Completed at Time of Every Nurse Visit

AR-3  V.N.A. Activity Record—ADL Completed at Time of Every Nurse Visit

AR-4  V.N.A. Activity Record—Planning & Coordination Completed at Time of Every Nurse Visit

AR-5  V.N.A. Activity Record—Psychological Therapy Completed at Time of Every Nurse Visit

 

 

lgee specific forms on pages which follow.
The terminal interview was made by the newly assigned observer.

LOT

3 Questions 8 and 9 were omitted at intake on A4.
A-5 was also completed at time of permanent move.

Figure 13—1. Chronological Schedule of Evaluations

SUOI}IUIJA(] PUB SULIO,| ‘9[NPAYDS
Effects of Continued Care

Study No.

Months of Follow

Date Due

 

Interview Schedule

Reference Date

 

Forms due (cross off when completed)

Form No. A-1

 

Admission

 

Post discharge

 

1st

quarter

 

2nd

quarter

 

3rd

quarter

 

4th

quarter

 

5th

quarter

 

6th

quarter

 

7th

quarter

 

8th

quarter

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Additional Forms

 

Form number

 

Dates due

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Patient's Name

Address Phone

 

Interviewer

 

108

Form No.

A-1
Schedule, Forms and Definitions

Admission Socioeconomic Evaluation

 

 

 

Name Study No.
Date of admission into study , Interviewer
Informant Relationship to patient

 

1. Birthday (month-day-year)

 

2. Sex (check): Male Female Color (check): White Non-white

Ww

Occupation
If (he) (she) worked during the past 10 years describe the job or business held last
week, if any, by answering the following questions. If no job or business last week,

give the information for the last job or business during the 10 year interval.

a. For whom did he work?

 

(Write name of company, business, organization or other
employer)

b. What kind of business or industry was this?
(examples: poultry hatchery, county junior
high school, auto assembly plant, radio
and TV service, retail supermarket, highway
construction)

 

c. What kind of work was he doing?
(examples: truck driver, 8th grade English
teacher, paint sprayer, repairs TV sets,
grocery checker, civil engineer)

 

d. Class of worker (check one)?

Employee of a private company, business or
individual for wages, salary, or commissions

 

Government employee (Federal, State, or local)

 

Self-employed in own business, professional
practice or farm

 

Working without pay in a family business or
farm

 

Other (specify)

 

Form No. A-2 109
Effects of Continued Care

Name

4.

110

Admission Socioeconomic Evaluation (continued)

 

Study No.

 

Circle the trade class that applies to usual occupation.
(Usual occupation is the kind of work patient has done in the past and would be doing
now except for age, illness, or injury.)

a. Higher executives, large proprietors (over $100,000), and major professionals.

b. Business managers (large concern), proprietors (medium businesses $35,000 to
$100,000), lesser professionals.

c. Administrative personnel, proprietors (small businesses $6,000 to $35,000),
semi-professionals, farm owners ($20,000 to $55,000).

d. Clerical and sales workers, technicians, proprietors (little businesses, $3,000 to
$6,000), farm owners ($10,000 to $20,000).

e. Skilled manual employees, small farm owners (under $10,000).
f. Machine operators and semi-skilled employees.
g. Unskilled employees, sharecroppers.

Education attainment (check appropriate column and circle highest grade completed).

()OK123456738 () College 1 2 3 4 Grad. Prof. Postgrad.
() HS. 12 34 (Indicate highest degree awarded )
( ) If foreign educated approximate ( ) Unknown (give best approx.)

equivalent in U.S.

Place of birth: In what state, U.S. possession, or foreign country was patient
born?

 

Date of last employment (or, if housewife, date when patient discontinued management
of her own home)

 

If patient is a widow or widower, what was the date of death of spouse?

 

Form No. A-2
Schedule, Forms and Definitions

Social Interaction

 

 

 

Name Study No.
Date of evaluation Interviewer
Informant Relationship to patient

 

Nature of evaluation (check one):

Admission 6 mo. 1 yr. 1 yr. 6 mo. 2 yrs.
Number of social contacts in or out of the home:
1. How many living children do you have?* daughter (s) son(s)_____

a. How often did you get together with them?

Daily

Twice or more weekly
Weekly

Every 2 to 3 weeks
Monthly

No regular visits
b. Other contacts with them (i.e., letter, phone, etc.)?

Specify:

 

2. Frequency of interaction with relatives (i.e., brothers, sisters, grandchildren,
in-laws, nephews, nieces).

Daily

Twice or more weekly
Weekly

Every 2 to 3 weeks
Monthly

No regular visits

3. Frequency of interaction with friends.

Daily

Twice or more weekly
Weekly

Every 2 to 3 weeks
Monthly

No regular visits

4. Frequency of interaction with neighbors.

Daily

Twice or more weekly
Weekly

Every 2 to 3 weeks
Monthly

No regular visits

*Indicate if respondent voluntarily reports deceased children:
Number Estimated date(s)

Form No. A-3 111
Effects of Continued Care

Social Interaction (continued)

Name Study No.

 

5. Now, how about the people you saw for special reasons (for example, the banker or
financial adviser, guardian, bus driver, mailman, salesman, newsboy, waiter, etc.) —
about how many of these did you see regularly?

Specify:

 

 

6. How many other social activities did you attend (i.e., church, clubs, etc.)?

Specify:

 

7. In the course of a day, you talked to about how many people? (including phone calls)

 

Is this more than, less than, or about the same as when you were younger, say for
example, when you were 45?

112 Form No. A-3
Schedule, Forms and Definitions

Social Interaction

(Terminal Evaluation)

 

 

 

Name Study No.
Date of evaluation Interviewer
Informant Relationship to patient

 

Number of social contacts in or out of the home:

 

1. How many living children do you have? daughter (s) son(s)

a. How often did you get together with them?____

Daily

Twice or more weekly
Weekly

Every 2 to 3 weeks
Monthly

No regular visits

b. Does the patient have any other contact (indirect) with the children such as by
letter, phone, other message, etc.?
(yes or no)

c. Where does the child live whom the patient sees most often?
With the patient
In the neighborhood
Outside of the neighborhood

 

2. Frequency of interaction with relatives (i.e., brothers, sisters, grandchildren,
in-laws, nephews, nieces).

Daily

Twice or more weekly
Weekly

Every 2 to 3 weeks
Monthly

No regular visits

3. Interaction with friends (single most frequent interaction).

Daily

Twice or more weekly
Weekly

Every 2 to 3 weeks
Monthly

No regular visits
Does this friend live-

In the neighborhood

Out of the neighborhood

Form No. T A-3 113
Effects of Continued Care

Social Interaction (continued)

(Terminal Evaluation)

 

 

Name Study No.
4, Interaction with friends (second most frequent interaction).

Daily

Twice or more weekly

Weekly

Every 2 to 3 weeks

Monthly

No regular visits
Does this friend live-

In the neighborhood

Out of the neighborhood ___

5. Now, how about the people you saw for special reasons (for example, the banker or
financial adviser, guardian, bus driver, mailman, salesman, newsboy, waiter, etc.) -

about how many of these did you see regularly?

Specify:

 

 

6. How many other social activities did you attend (i.e., church, clubs, etc.)?

Specify:

 

7. In the course of a day, you talked to about how many people? (including phone calls)

 

Is this more than, less than, or about the same as when you were younger, say for
example, when you were 45?

114 Form No. T A-3
Schedule, Forms and Definitions

Social and Economic Function

 

 

 

Name Study No.
Date of evaluation Interviewer
Informant Relationship to patient

 

Nature of evaluation (check one):

 

Admission 3 mo. 6 mo. 9 mo. 1 yr.
Post Discharge 1 yr. 3 mo. 1 yr. 6 mo. 1 yr. 9 mo. 2 yrs.
1. Residence address (complete mailing, including postal zone):

6.

 

 

 

Phone No.
Nature of residence (check one):
Single home Hotel Nursing
Double home Boarding Hospital
Apartment Custodial Other (specify)

Is patient legally designated owner of his residence (check):
Yes No

Is patient gainfully employed 1 or more hours per week (yes or no)?
(Including even such employment as baby sitting or other employment in the home.)

If "yes" indicate nature and hours per week of employment

 

If unemployed since last visit, indicate date

 

Check the sources of funds used for food, clothing, shelter, and medical care:

 

Private industrial

. Aid for the aged, plans (retirement
Parents, children, . . .
1g . social agencies, Social or insurance), .
siblings, their . . Other (specify)
rogeny or Spouses state, city, or Security | savings, salary,
progeny P county stocks or bonds

(pt. or spouse)

 

 

 

 

 

 

 

 

Marital status (check): Married Never married Divorced

Separated _ Widow(er)

If spouse deceased since last visit, specify date

 

Form No. A-4 115
Effects of Continued Care

Social and Economic Function (continued)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Name Study No.
7. Household composition:
. . Indicate specific kinds of medical
Relationship
Age Sex housekeeping, or supervisory
to patient . .
assistance given to patient
Head—=
Notice:

Omit the next two items from the Intake Interview only.

8. Has anyone close to you passed away since I talked with you (i.e., children, grand-
children, friends and neighbors, other relatives, and service personnel)?

No Yes

If "yes,'

' who?

 

9. Has anyone close to you moved since I talked with you, so that the inconvenience of
travel prevents you from getting together with them now?

No Yes

116

If "yes," who?

 

Form No. A-4

 
Schedule, Forms and Definitions

Social and Economic Function

 

(Terminal Evaluation)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Name Study No.
Date of evaluation Interviewer
Informant Relationship to patient
1. Residence address (complete mailing, including postal zone)
Phone No.

2. Nature of residence (check one):

Single home Hotel Nursing

Double home Boarding Hospital

Apartment Custodial Other (specify)
3. Is patient legally designated owner of his residence (check):

Yes No .

4, 1s patient gainfully employed 1 or more hours per week (yes or no)?

(Including even such employment as baby sitting or other employment in the home and

consider employment as activity which is functionally involved in procuring income.)

a. If "yes" indicate nature and hours per week of employment

b. Indicate date patient retired from employment

c. Indicate date patient ceased being home maker
5. Check the sources of funds used for food, clothing, shelter, and medical care:

Private industrial
Parents, children, aid for the aged, } plans (retirement
. social agencies, Social or insurance), .
siblings, their . . Other (specify)
ony or Spouses state, city, or Security | savings, salary,
progeny pou county stocks or bonds
(pt. or spouse)
6. Marital status (check): Married Never married Divorced
Separated Widow(er)

If spouse deceased, specify date

 

Form No. T A-4 117
Effects of Continued Care

Name

Social and Economic Function (continued)

(Terminal Evaluation)

Study No.

 

7. Household composition:

 

Relationship
to patient

Age

Sex

Indicate specific kinds of medical
housekeeping, or supervisory
assistance given to patient

 

Head—=

 

 

 

 

 

 

 

 

 

 

 

 

 

118

Form No. T A-4
Schedule, Forms and Definitions

Activities of Daily Living

 

 

 

 

Name Study No.
Date of evaluation Interviewer
Informant Relationship to patient

 

Nature of evaluation (check one):
Admission 3 mo. 6 mo. 9 mo. 1 yr.
Post discharge 1 yr. 3 mo. 1 yr. 6 mo. 1 yr. 9 mo. 2 yrs.

For each area of functioning listed below, check description that applies. (The word
"assistance" means supervision, direction or personal assistance.)

 

 

Bathing — either sponge bath, tub bath, or shower —

- C]
Receives no assistance Receives assistance in Receives assistance in
(gets in and out of bathing only one part bathing more than one
tub by self if tub is of the body (such as part of the body (or
usual means of bathing) back or a leg) not bathed)

 

Dressing — gets clothes from closets and drawers — including underclothes, outer garments
and using fasteners (including braces if worn) —

CO C] C
Gets clothes and gets Gets clothes and gets Receives assistance in
completely dressed dressed without getting clothes or in
without assistance assistance except getting dressed, or
for assistance in stays partly or
tying shoes completely undressed

 

Toileting — going to the "toilet room" for bowel and urine elimination; cleaning self
after elimination, and arranging clothes —

CO] CO] C]
Goes to ''toilet room," Receives assistance in Doesn't go to room
cleans self, and arranges going to "toilet room" termed "toilet" for
clothes without assistance or in cleansing self the elimination process
(may use object for support or in arranging clothes
such as cane, walker, or after elimination or in
wheelchair and may manage use of night bedpan
night bedpan or commode, or commode

emptying same in AM)

 

Transfer —

1 2 3
Moves in and out of bed Moves in or out of bed Doesn't get out of bed
as well as in and out of or chair with assistance

chair without assistance
(may be using object for
support such as cane or

walker)

 

Form No. A-5 119
Effects of Continued Care

Name

Activities of Daily Living (continued)

Study No.

 

For each area of functioning listed below, check description that applies. (The word

"assistance" means supervision, direction or personal assistance.)

 

Continence —

1
Controls urination and
bowel movement completely
by self

2

Has occasional "accidents"

C3
Supervision helps keep
urine or bowel control;
catheter is used, or is
incontinent

 

Feeding —
1

[]

Feeds self without
assistance

2

Feeds self except for
getting assistance in
cutting meat or buttering
bread

=
Receives assistance in
feeding or is fed partly
or completely by using
tubes or I.V. fluids

 

Walking —

1
Walks on level without
assistance

=
Walks without assistance
but uses single, straight
cane

C]
Walks without assistance
but uses two points for
mechanical support such
as crutches, a walker or
two canes (or wears a
brace)

4

Walks with assistance
5

Uses wheelchair only

6

Not walking or using
wheelchair

 

House-conf inement —
1

Has been outside of resi-
dence on 3 or more days
during past 2 weeks

2

Has been outside of resi-
ence on only 1 or 2 days
during past 2 weeks

3

Has not been outside of
residence in past 2
weeks

 

Specify the person or persons giving assistance in each dependent function:

1

3

 

Spouse

Parents, children,
siblings, or
progeny or spouse
of any

nurse

Attendants, companions,
housekeepers, institutional
staff, friends, visiting

Other (specify)

 

Bathing

 

Dressing

 

Going to toilet

 

Transfer

 

Continence

 

Feeding

 

 

 

Walking

 

 

 

 

120

Form No. A-5
Schedule, Forms and Definitions

Medical Care

 

 

 

Name Study No.
Date of evaluation Interviewer
Informant Relationship to patient

 

Nature of evaluation (check one):

 

Post Discharge 3 mo. 6 mo. 9 mo. 1 yr.

 

1 yr. 3 mo. 1 yr. 6 mo. 1 yr. 9 mo. 2 yrs.

1. a. Since last interview has the patient been in a hospital, nursing home, or other
custodial residence? Yes No

b. If "yes," list:
Nursing Home or Admission Discharge
Hospital Custodial Residence Date Date
(mo.-day-yr.) (mo.-day-yr.)

 

 

 

 

 

2. a. Does the patient have hospital insurance? Yes No

b. If "yes," enter the name of the company

 

3. a. Since last interview, has the patient made a permanent change in residence?
Yes No

 

b. If "yes," also complete both forms A-5 and A-7 for every permanent move during the
interval.

4.% a. Last week or the week before, did a doctor examine or talk to the patient?
Yes No (if "no" skip to Question 5).

 

 

b. How many times during the past 2 weeks?

 

c. Where did the patient talk with the doctor?

Home Dr.'s Office Hospital Clinic Hospital Telephone

 

 

 

 

 

 

 

 

Times

 

5.% a, How long has it been since a doctor last examined or talked to the patient?
weeks months

b. Where did the contact take place?

Home Dr.'s Office Hospital Clinic Hospital Telephone

 

 

 

 

 

 

 

 

*Questions 4 through 7 refer to direct contacts. Indirect
contacts should be described in REMARKS.

Form No. A-6 121
Effects of Continued Care

Name

Medical Care (continued)

Study No.

 

5.% c.

6.% a.

7.% a.

Who is the patient's regular doctor? Name

 

Address

 

Last week, or the week before, did a nurse have professional contact with the
patient? Yes No (if "no," skip to Question 7)

How many times during the past 2 weeks?

 

What kind of nurses were they?

Visiting Nurse County Nurse Practical Nurse Other (specify)

 

Times

 

 

 

 

 

 

 

How long has it been since a nurse last had professional contact with the patient?
weeks months

What kind of nurse was she?

Visiting Nurse County Nurse Practical Nurse Other (specify)

 

 

 

 

 

 

 

 

8. Since the last interview, has the patient received service from:

Dentist, number of visits

 

Optometrist, number of visits
Chiropodist, number of visits
Social worker, number of visits

Physical therapist, number of visits

*Questions 4 through 7 refer to direct contacts. Indirect contacts should be described in
remarks.

REMARKS :

122

Form No. A-6
Name

Schedule, Forms and Definitions

Medical Care

(Terminal Evaluation)

Study No.

 

Date of evaluation

Interviewer

 

 

Relationship to patient

 

Informant
1. a. Since last
custodial r
b. If "yes," 1
Hos

interview has the patient been in a hospital, nursing home, or other

 

 

esidence? Yes No
ist:

Nursing home or Admission Discharge
pital Custodial Residence Date Date

 

(mo.-day-yr.) (mo.-day-yr.)

 

 

 

 

4o% a,

Last week or the week before, did a doctor examine or talk to the patient?
Yes No (if "no" skip to Question 5)

b. How many times during the past 2 weeks?

 

c. Where did the patient talk with the doctor?

Home

Dr.'s Office

Hospital Clinic

Hospital

Telephone

 

Times

 

 

 

 

 

 

weeks

 

b. Where did the contact take place?

Home

Dr.'s Office

months

 

Hospital Clinic

Hospital

How long has it been since a doctor last examined or talked to the patient?

Telephone

 

 

 

 

 

 

c. Who is the patient's regular doctor?

Yes

Name

 

Address

 

Last week, or the week before, did a nurse have professional contact with the

 

 

patient?

b. How many times during the past 2 weeks?

c. What kind of nurses were they?

Visiting Nurse

No (if "no,"

County Nurse

skip to Question 7)

 

Practical Nurse

Other (specify)

 

 

 

Times

 

 

 

*Refers to direct contacts.

Form No. T A-6

Indirect contacts should be described in remarks.

123
Effects of Continued Care

Medical Care (continued)

(Terminal Evaluation)

 

 

Name Study No.
7.*% a. How long has it been since a nurse last had professional contact with the patient?
weeks months
b. What kind of nurse was she?
Visiting Nurse County Nurse Practical Nurse Other (specify)

 

 

 

 

 

8. During the last 2 weeks has the patient received service from
Dentist, number of visits

Optometrist, number of visits

 

Podiatrist, number of visits

 

Social worker, number of visits

Physical therapist, number of visits

9. In the past 3 months did the patient receive service? If so
Dentist, number of visits
Optometrist, number of visits

Podiatrist, number of visits

 

Social worker, number of visits
Physical therapist, number of visits
10. a. Did a nurse take care of this patient in the home between discharge from Abington

House (formerly Benjamin Rose Hospital) and this evaluation (terminal and/or
death)? Yes No

 

 

If "yes," specify:

b. Did any other person come into the home to help with the care of the patient be-
tween discharge from Abington House (formerly Benjamin Rose Hospital) and this
evaluation (terminal and/or death)? Yes No

 

' circle:

If "yes,'
Relative — friend/neighbor — homemaker — part-time sitters

Others

 

*Questions 4 through 7 refer to direct contacts. Indirect contacts should be described in
REMARKS.

REMARKS :

124 Form No. T A-6
Name

Schedule, Forms and Definitions

Permanent Move

Study No.

 

Date of evaluation

Interviewer

 

Informant

 

Relationship to patient

 

Nature of evaluation (check one):

 

3 mo.
1 yr. 3 mo.

 

 

1 yr.

6 mo.
6 mo.

 

9 mo.
9 mo.

1 yr.
2 yrs.

 

1 yr.

 

1. Date of permanent move (mo.-day-yr.)

2. Address of move (complete mailing, including postal zone)

 

 

 

3. Nature of residence (check one):

 

 

Single home Hotel Nursing
Double home Boarding Hospital
Apartment Custodial Other (specify)
4. Household composition:
Indicate specific kinds of medical,
Relationship Age Sex housekeeping, or supervisory

assistance given to patient

 

Head—=

 

 

 

 

 

 

 

 

 

 

 

 

 

 

5. Did you move because you were sick or needed care?

Form No. A-7

(yes or no)

125
Effects of Continued Care

Name

Physical Examination

Study No.

 

Date of evaluation

Nature of evaluation (check one):

Physician

 

History (check):

Present

Absent Unknown

1. [1] [] ]

2.

 

 

 

2.

 

 

 

 

 

 

7.

8.

 

Entry Terminal

 

Chest pain for which an M.D. advised bed rest for 2 or
more weeks (or history of diagnosis of coronary at-
tack).

Distressful sensation relieved within a few minutes by
nitroglycerin or rest
(occurring in precordium, any part of either upper
extremity, left scapular region, angles of jaw,
chin, or roof of mouth
&
precipitated by effort, acute emotional disturb-
ance, a large meal, exposure to cold, or walking
against the wind).

Non-inflammatory polyarthritis (motion limited by pain
in 2 or more joints) during past month.

Inflammatory polyarthritis during past month (pain and
redness or swelling in 2 or more joints).

Any diastolic murmur in the mitral or aortic area in
the absence of known congenital or syphilitic heart
disease.

A Grade 3 or more mitral or aortic systolic murmur
(Grade 1 is barely audible and sometimes not heard at
first; Grade 2 is softly audible immediately).

Record blood pressure in left arm: start of exam

Current treatment (check):

 

9.

10.

126

Yes No

 

 

 

 

 

 

J]

end of exam

9. Digitalis or other cardiac glycoside.

10. Insulin or drug to reduce blood sugar.

Form No. A-10
Schedule, Forms and Definitions

Physical Examination (continued)

Range of Movement and Strength

 

 

 

 

 

 

Name Study No.
Right Left ]
Anatomi-— Does Does
cal Test Instruction Does Does
Area not Without [Against |N.T.* not Without |Against|N.T.*
Resist-|Resist- Resist-|Resist-
ance ance ance ance

 

Shoulder| Raise hand forward high
above head

 

Put hand on back of head

 

Swing arm toward and away
from midline at shoulder
level

 

Elbow Bring hand to front of
shoulder

 

With elbow at shoulder
level straighten out arm

 

Wrist Wave hand side to side

 

Fingers | Grasp small object with
fingers

 

Extend fingers

 

Thumb Make tip of thumb touch
tip of little finger

 

After touching fingertip
extend thumb

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Knee Kick foot into air +
Hip Bring knee toward chest
Bring leg out to side
Return leg to center
Ankle Move foot from side to
side
Bend foot up and down
*N.T. = Not Tested. Indicate reason whenever patient is not tested.

Form No. A-10 127
Effects of Continued Care

Physical Examination

(Terminal Evaluation)

Name Study No.

Date of Evaluation Physician

 

 

History (check):

 

Present Absent Unknown

3. [1] 3. Non-inflammatory polyarthritis (motion limited by pain
in 2 or more joints) during past month.

| [] "4, Inflammatory polyarthritis during past month (pain and

redness or swelling in 2 or more joints).

4.

7. Record blood pressure in left arm: start of exam
end of exam

11. List each "fracture" or "broken bone" that patient has sustained during the past two
years.

 

 

 

 

 

 

128 Form No. T A-10
Physical Examination (continued)

(Terminal Evaluation)

Range of Movement and Strength

 

Schedule, Forms and Definitions

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Name Study No.
Right Left
Anatomi- Does Does
cal Test Instruction Does Does
Area hot Without |Against|N.T.* nok Without [Against |N.T.*
Resist-|Resist- Resist-|Resist-
ance ance ance ance
Shoulder |Raise hand forward high
above head
Put hand on back of head
Swing arm toward and away
from midline at shoulder
level
Elbow Bring hand to front of
shoulder
With elbow at shoulder
level straighten out arm
Wrist Wave hand side to side
Fingers |Grasp small object with
fingers
Extend fingers
Thumb Make tip of thumb touch
tip of little finger
After touching fingertip
extend thumb
Knee Kick foot into air
Hip Bring knee toward chest
Bring leg out to side
Return leg to center
Ankle Move foot from side to
side
Bend foot up and down
*N.T. = Not Tested. Indicate reason whenever patient is not tested.
Form No. T A-10 129
Effects of Continued Care

Name Study No.
Nature of evaluation (check one): Entry Terminal
Date of test
WEIGHT (in indoor clothing — doesn't include coats or suit
coats — includes shoes)

HEIGHT (with shoes)

URINE PROTEIN (tested with Uristix)

URINE SUGAR (tested with Uristix)

BLOOD SUGAR (mg./100 cc.)

TIME BLOOD DRAWN (may use designation 2 hr. p.c.)

TIME LAST FOOD

NEARPOINT "E" CHART VISION SCREENING

C/T RATIO: (X-ray No. )

BUN (mg./100 ml.)
130 Form No.

Laboratory Examinations

 

 

A-11
Patient's Name

Physician

EKG Findings

Schedule, Forms and Definitions

Study No.

 

 

EKG FINDINGS

 

 

 

 

Date of EKG (mo.-day-yr.)

Q of 0.04 seconds or more in AVL

Q of 0.04 seconds or more in AVF

Qs
Qs
Qs

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Form No.

A-11

QS
Qr
Qr
Qr
Qr
QR
QR
QR
QR
QR

in Vs,

<
~

in
in
in
in
in
in
in
in
in

in

< dd dad << <<
wm Aw ey EW oy Wn

in

aN

greater than 25% in AVL

QR greater than 257 in AVF

Primary T

Primary

Primary

Primary

T
T
Primary T
T
T

Primary

changes
changes
changes
changes
changes

changes

in
in
in
in
in

in

Atrial fibrillation

lead I
lead AVR
lead V
lead V
lead V

[

lead V

Complete bundle branch block

Second degree atrio-ventricular block

Third degree atrio-ventricular block

131
Effects of Continued Care

Principal Diagnosis Leading to Admission

 

Name Study No.

 

Principal Diagnosis Leading to Admission
(identify only one for each case)

 

arteriosclerotic heart disease (including myocardial infarction)
cerebral infarction

chronic brain syndrome

depressive, anxiety or psychoneurotic reaction
diabetes mellitus

exogenous obesity

fracture of femur

generalized arteriosclerosis

hypertensive cardiovascular disease

malignancy

osteoarthritis

paralysis agitans

pneumonia

rheumatoid arthritis

urinary tract infection

other (specify)

 

Was malignancy present (within 5 years before admission)? Yes No

Did this patient sign out "against advice?" Yes No

132 Form No. A-12
Patient's Name

Study No.

Date of Evaluation

Nature of Evaluation

Form No.

A-20

Psychosocial Test

Q-Sort Distribution

 

Schedule, Forms and Definitions

 

 

 

 

 

 

Location

Examiner
(check one): Entry Terminal
1 2 3 4

 

133
Effects of Continued Care

Coloured Progressive Matrices Test (Raven)

Scoring Form

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Name Sex Age
School Grade
Test Begun Test Ended Total Time
A Ab B
1 1 1
2 2 2
3 3 3
4 4 4
5 5 5
6 6 6
7 7 7
8 8 8
9 9 9
10 10 10
11 11 11
12 12 12
Total Score
Percentile
134 Form No. A-21
Schedule, Forms and Definitions

Orientation and Mental Control Test

 

 

 

 

 

Patient's Name Location

Study No. Examiner

Date of evaluation Nature of evaluation (check one):
Entry Terminal

(Note: Take down patient's answers)

ORIENTATION FOR TIME

1. What year is it now? (Scored 1 or 0)
2. What month? (Scored 1 or 0)
3. What day? (Day of the week is OK) (Scored 1 or 0)

Score

ORIENTATION FOR PLACE

1. What is this place we are in now? (Abington House, formerly Benjamin
Rose = 2, Hospital = 1, other = 0)
(My home = 2, if at home)

2. What city are we in now? (Scored 1 or 0)
3. What is your address (at home)? (Scored 1 or 0)
Score
ORIENTATION FOR PERSON
1. What is your full name? (First & last = 2, first or last = 1,
other = 0)
2. How old are you? (Scored 1 or 0)
3. Are you married? How many children? (Both correct = 2, one correct = 1,
other = 0)
Score

MENTAL CONTROL

1. Say the alphabet for me, like this
"A, B, Cy, ...." (Within 30" no errors = 2, 30" one or
two errors = 1, other = 0)

2. Count backwards from 20 all the way to 1,
like this '20, 19, 18, ...." (30" no errors = 2, 30" one or two
errors = 1, other = 0)

3. Count by threes starting with 1 until I tell
you to stop. Like this, "1, 4, 7, .... add

three each time." (45" up to 40 with no more than 1
error = 2, with 2 errors = 1,
other = 0)
Score

Note if patient is untestable (give reason):

Total Orientation and Mental Control
SCORE (Maximum = 18)

Form No. A-22 135
Effects of Continued Care

Data from Death Certificate

Name Study No.

 

Information from death certificate (if deceased):
18. CAUSE OF DEATH

Part I. Death was caused by:
immediate cause (a)

 

Conditions, if any which Due to (b)
gave rise to above cause
(a), stating the under-
lying cause last. Due to (ec)

 

 

 

Part II. Other significant conditions contributing to death but not related to
the terminal disease condition given in Part I (a)

 

Date of Month Day Year 19. Was autopsy performed?
Death: Yes No

 

 

Did death occur in hospital? (Yes or No)

 

"

Name of hospital, if "yes"

 

Did an M.D. attend the deceased on the day of death (Yes or No)

136 Form No. A-18
Schedule, Forms and Definitions

Socio-Medical History of Expired Patient

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Name Study No.

Date of evaluation Interviewer

Informant Relationship to patient

1. Indicate date patient retired from employment

2. Indicate date patient ceased being homemaker

3. If spouse deceased, specify date

4. a. Since last interview has the patient been in a hospital, nursing home, or other
custodial residence? Yes_______ No

b. If "yes," list:

Nursing Home or Admission Discharge
Hospital Custodial Residence Date Date
(mo.-day-yr.) (mo.-day-yr.)
5. a. Did a nurse take care of this patient in the home between discharge from Abington
House (formerly Benjamin Rose Hospital) and this evaluation? Yes ______ No
If "yes," specify:

b. Did any other person come into the home to help with the care of the patient be-
tween discharge from Abington House (formerly Benjamin Rose Hospital) and this
evaluation? Yes No
If "yes," circle:

Relative — friend/neighbor — homemaker — part-time sitters
Others
6. List each "fracture" or "broken bone' that patient sustained during the past two

Form Nos. T A-4, T A-6, T A-10

years.

 

 

 

137
Effects of Continued Care

Greater Cleveland Confidential Referral Form

 

Hospitals & Physicians and Public Health Nursing & Other Health Agencies

Use form to refer for continuity of care even if telephone contact is made. Complete in
triplicate. Mail two copies to designated agency. One copy will be returned with a
report for patient's record.

 

 

To: Patient: Hosp. No.
0.P.D. No.
Agency No.

(Address) (Name)

Date of referral: Tel. No.

 

(Address in detail)

 

 

 

 

 

From: Birth Date: Sex: S MWD Sep.
(Address) (Name of nearest relative or friend)
Tel. No. Ext. Hosp. Adm. Date: ____ ~~ Disch. Date:
Next App't. Date: Clinic
(Person to contact) (Name)
Physician's Diagnosis & Prognosis: Is patient aware of diagnosis & prognosis?

Is family aware of diagnosis & prognosis?

Orders including Medication & Diet Instruction:

Date: Physician's Signature:

 

 

 

Comments by Hospital Nurse:

Nurse's Signature:

 

138 E 17 - 66
Schedule, Forms and Definitions

Greater Cleveland Confidential Referral Form (continued)

Comments by Dietitian, Social Worker, Therapist (Occupational, Physical, Speech), Other:

Dates: Signatures:

 

 

 

Comments by Public Health Nurse and/or Other Worker in Community Agency:

Agency: Signatures:

 

Dates:
Address: ~~ === Tel. No.___ ~~ Ext.

 

 

Authorization to Release Information:

I hereby authorize to release such
(Name of Clinic, Hospital and/or Physician)
information in connection with this form as it, in its sole discretion, may deem proper.

 

Date: Signature:

 

(Patient or Legal Representative — Guardian if Minor)

E 17 - 66 139
Effects of Continued Care

V.N.A. Monthly Activity Record — Exercises

Patient's Name

Address

Month and Year

 

 

 

Number of Visits this Month

 

EXERCISES

List specific exercises

for individual joints:

number of repetitions
and frequency; method
in which passive

motion is carried out;

etc.

Remarks:

 

Assisted
Patient

Tested or
Observed
Performance

Instructed
Patient
Verbal/Demons.

Instructed
Family Member
or Other

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Nurse or Physical Therapist

140

 

Form AR-1
Schedule, Forms and Definitions

V.N.A. Monthly Activity Record — Prescribed Treatment

 

Patient's Name Month and Year

 

 

Remarks
PRESCRIBED TREATMENT

Medications; Dressings;
Treatments; Orthopedic
Equipment
(List as Prescribed)

 

Instructed
Family Member
or Other

Reviewed Instructed

Administered Procedure Patient

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Nurse or Physical Therapist

 

Form AR-2 141
Effects of Continued Care

V.N.A. Monthly Activity Record — ADL

Patient's Name Month and Year

 

Remarks:

ACTIVITIES OF
DAILY LIVING

 

 

 

Assisted Tested or Instructed Instructed
Patient Family Member
Observed Patient
or Gave Care or Other
Bathing
Dressing

 

Going to Toilet and
Toilet Training

 

Transfer

 

Feeding

 

Walking

 

Stair Climbing

 

Wheelchair Management

 

Other (Specify)

 

Other

 

Other

 

 

 

 

 

 

 

Nurse or Physical Therapist

142 Form AR-3
Schedule, Forms and Definitions

V.N.A. Monthly Activity Record — Planning and Coordination

Patient's Name

Month of Follow

Date Due

Month and Year

 

 

HEALTH
PROMOTION

Remarks:

 

Listened with
Appropriate Response

Identified
Problem

Interpreted
and/or Instructed

 

Patient Family

Patient Family

Patient Family

 

Nutrition

 

Accident Prevention

 

Medical Supervision

 

Personal Hygiene

 

Home Management

 

Recreation

 

Other (Specify)

 

 

 

 

 

 

 

 

PLANNING AND
COORDINATION

(Indicate individual
or group with whom
you are working.)

Remarks:

 

Identified
Problems

Transmitted
Reports

Developed
Plan

Made
Referral

 

Patient and Family

 

Physician

 

Other (Specify
individual or group)

1.

 

2.

 

 

 

 

 

 

 

 

 

 

Nurse

 

Form AR-4

143
Effects of Continued Care

V.N.A. Monthly Activity Record — Psychological Therapy

 

Patient's Name Month and Year

 

 

Remarks:

PSYCHOLOGICAL
THERAPY

 

Listened with

Appropriate Response Interpreted Proposed Action

 

Patient Family Patient Family Patient Family

 

Reaction to Illness

 

Disease Process and
Prognosis

 

Social and Economic
Problems

 

Limitations Imposed by
Illness:

 

Physical

 

Economical

 

Social

 

Feelings about the
Treatment Plan

 

Interpersonal Relations
with the Family

 

Other (Specify)

 

Other

 

Other

 

 

 

 

 

 

 

 

 

Nurse

 

144 Form AR-5
Schedule, Forms and Definitions

SUPPLEMENTARY DEFINITIONS

The following information supplements informa-
tion about measures described in Chapter 5. For
measures that are not described adequately in
published literature, definitions are expanded to
aid the research observer in completing the
preceding forms.

Study Form A-2.

The occupation of a woman’s husband is ordi-
narily the occupation factor used in defining
socioeconomic class by Hollingshead’s Two
Factor Index of Social Position; however, it is
not appropriate to use the husband’s occupation
for certain aged, long-retired, and long-widowed
people. After consulting Dr. Hollingshead, the
following modified definitions have been ac-
cepted for this study. Unless otherwise defined
herein, the occupation used in calculating the
Index of Social Position is that of the male head
of the household, even if retired. Women are,
thus, classified according to their husband’s
highest occupational level as in the case of
housewives, widows who have never worked,
divorcees who have never worked, and women
who have been widowed for less than ten years.
If unmarried, widowed for more than ten years
and working, widowed for more than ten years
and retired, divorced and working, or divorced
and retired, the woman’s usual occupation is
used in calculating the Index position.

Study Form A-4.

The Index of Independence in Socioeconomic
Functioning was developed to measure socioeco-
nomic function of the aged (The Staff of The
Benjamin Rose Hospital: Multidisciplinary
Studies of Illness in Aged Persons. A New
Classification of Socioeconomic Functioning of
the Aged. J. Chron. Dis. 13: 453, 1961). A
modified form, namely, the Index of Economic
Dependence, is used in the present study and
incorporates indicators of resources and pro-
ductivity as defined below.

The observer records the economic status of
patients as it existed during the 2-week period
which preceded the due date of the evaluation.

At the time of the initial research evaluation,
the period of evaluation is defined as the 2-week
period which preceded the current period of
continuous hospitalization. As a guide to the
interviewer who is responsible for completing
Study Form A-4, additional details of definition
are presented here concerning the terms gainful
employment, home ownership, and agency sup-
port.

Patients are considered gainfully employed if
they derive income from an activity which they
perform regularly at least one hour each week,
including even such activities as baby sitting or
active involvement in managing a boarding
home.

In general, home ownership is identifiable
precisely; however, certain clarifications of defi-
nition assure comparable recording for use in the
Index. Since the concern here is with functional
residence, a patient who lives in a residence
which he does not own is classed as “not” the
home owner though he may be the legally
designated owner of some other dwelling, while
a patient who lives in a residence that he or his
spouse owns is considered to be the home owner
whether he or someone else pays the property
taxes. A patient who is co-owner of his place of
residence is, also, classed as a home owner.
Trailer ownership is not defined as home owner-
ship unless the patient or his spouse own the
land or pay taxes on the land on which the
trailer stands.

In order to determine the presence or absence
of agency support, the observer reviews the
sources of funds used for the patient’s food,
clothing, residence tax or rent, medicines, and

doctor’s fees. Though all of the information
gathered in this question is not needed in order

to determine the presence or absence of agency
support, the detailed review assures the best
possible validity and reliability of the measure.
An agency is defined as any social service or
charitable agency, local, county, state, or na-
tional, private or public, which assigns funds to
the patient or his spouse for the aforementioned
needs. Social Security, veterans’ pensions,

145
Effects of Continued Care

unemployment compensation, family assistance,
rental from properties, and retirement plans
such as industrial and teachers’ retirement plans
are not classified as sources of agency support in
the Index of Economic Dependence. For the
purposes of this measure, food coupons or
certificates are defined as funds, while free
services or equipment are not so defined. Pay-
ment by another of a patient’s monetary obliga-
tions are considered as use of “funds” though
the money may not be given directly to the
patient.

Based on the above definitions, the following
four grades are included in the Index of Eco-
nomic Dependence in order of increasing eco-
nomic dependence. Grade I includes people who
own their residence, are employed, and do not
receive agency support. Grade 2 includes people
who are either unemployed or do not own their
residence (not both), and do not receive agency
support. Grade 3 includes people who either
receive agency support or do not own their
residence (not both), and are unemployed.
Grade 4 includes people who are unemployed,
do not own their residence, and receive agency
support.

Study Form A-5.

Evaluations of bathing, dressing, going to toilet,
transferring, continence, and feeding comprise
the measures of self-function in individual activi-
ties of daily living. Functional status in these
activities have been summarized in a standard-
ized measure, the Index of ADL, which reflects
the adequacy of organized neurological and
locomotor response and which serves as a graded
measure of such biological and psychosocial
function (Katz, S., Ford, A.B., Moskowitz,
R.W., Jackson, B.A. and Jaffe, M.W.: Studies of
Illness in the Aged, The Index of ADL: a
Standardized Measure of Biological and Psycho-
social Function. J. A.M.A. 185:914, 1963).

By means of a series of questions and observa-
tions, the observer forms a mental picture of the
patient’s ADL status as it existed during the
2-week period which preceded the due date of
the evaluation. The observer determines whether
another person assisted the patient or whether
the patient functioned alone, defining assistance

146

as active personal assistance, directive assistance,
or supervision. The actual existence of such

assistance is considered in the evaluation, not
the potential or ability of the patient. Thus, for

example, overprotective assistance is defined as
assistance though the observer considers the
patient as more able, and refusal to perform a
function is defined as nonfunctioning though
the patient is deemed able. The observer uses the
following definitions in completing Study Form
A-5 and, in general, records the most dependent
degree of function which was performed during
the 2-week period.

Bathing is the overall complex behavior of
getting water and cleansing the whole body. A
patient receives ‘“‘no assistance” (first of the
three classes of bathing on Study Form A-5) if
no other person is involved in any part of the
process of taking a sponge, shower, or tub bath
to wash the whole body. Such a patient goes to
the sink by himself if he sponge-bathes at the
sink, gets in and out of a tub by himself if he
tub-bathes, and is not supervised in the shower if
showering is the means of bathing. A patient
receives ‘‘assistance in bathing only one part of
the body” if he functions by himself as defined
above, except that he is assisted in washing only
one part of the body, as his back alone or one
leg alone. The class “assistance in bathing more
than one part of the body” includes the individ-
ual who is assisted in washing more than one
part of the body or who does not bathe. This
last, most dependent category, includes also the
following: the patient to whom water is brought
even though he washes himself; the patient who
is taken to the place of bathing though he
washes himself; the person who is helped in or
out of a tub as regularly as once a week; the
patient who is regularly supervised for reasons of
safety though he washes himself; and the patient
who can’t reach his feet to wash them.

Dressing is the overall complex behavior of
getting clothes from closets and drawers and
then getting dressed. A patient gets “completely
dressed without assistance’ (first of the three
classes of dressing on Study Form A-5) if no
other person is involved in getting clothes from
closets and drawers nor in putting on the
clothes, including brace, if worn, and including
outer garments and footwear. Fasteners must
also be managed without assistance. Footwear
include such items as socks and slippers or shoes.
The intermediate category of dressing on Study
Form A-5 includes those who get their own
clothes and dress independently as defined
above “except for assistance in tying shoes.” A
patient is placed in the third and most depend-
ent category if he receives ‘“‘assistance in getting
clothes or in getting dressed’ or remains “partly
or completely undressed.”

Going to toilet is the act of going to the room
termed the “toilet room” for bowel and bladder
function, transferring on and off the toilet,
cleaning after elimination, and arranging clothes.
The patient who functions wholly by himself,
including getting to the room, is classed as
functioning ‘without assistance’ (first of the
three classes of toileting on Study Form A-5). It
should be noted that an individual in this class
may or may not be using an object for support
such as a cane, walker, or wheelchair; and he
may be using a bedpan or commode at night, in
which case he empties it himself to be consid-
ered in the ‘without assistance” category. If
another person assists in any part of the func-
tion, the toileting status is recorded as “receives
assistance” (intermediate toileting category on
Study Form A-5). Toileting status is also re-
corded as “receives assistance’ for an individual
who uses the toilet room at certain times, and at
other times uses a daytime bedpan or commode.
The third category, namely, “doesn’t go to room
termed toilet”, is self-explanatory. Note that
toileting is not concerned with continence. A
patient is occasionally incontinent, but manages
himself completely independently insofar as
toileting is concerned. Toileting function, for
this patient, is recorded as “without assistance.”

Transfer is the process of moving in and out
of bed and in and out of a chair. If no other
person is involved in the transfer, the patient is
considered to function “without assistance”
(first of three classes of transfer on Study Form
A-5). Such a patient may be using an object for
support, e.g., cane, walker, or bedpost. The
intermediate category, namely, ‘with assist-
ance”, applies if another person is involved in
the process. Patients in the third category are

Schedule, Forms and Definitions

bedridden and do not leave the bed at all. In
evaluating transfer status, the observer may be
told that the patient is not allowed to transfer
unless supervised for reasons of safety. The
observer then tries to observe whether such
supervision is a reality. The observer may occa-
sionally find, for example, that a daughter
claims she supervises her mother whenever her
mother moves from one place to another, while
observation reveals that the mother moves about
entirely on her own, and the daughter actually
means that she is always within hearing distance.

Continence refers to the physiologic process
of elimination from bladder and bowel, where
incontinence is the involuntary loss of urine
and/or feces. The function is thought of as the
primitive function of control and does not
include any consideration of hygiene, toileting,
or constipation. The patient is classed as “‘con-
trols urination and bowel movement completely
by self” (first of three continence categories on
Study Form A-5) if no other person assists. Such
a patient can exert some degree of control on
the process himself by medication or by self-
administered enema (or, in the case of patient
with a colostomy, may manage this by himself).
A slight amount of wetness or slight soiling of
underclothes is occasionally noted by others and
is not perceived as incontinence by the patient,
in which case status is recorded as “controls
urination and bowel movement.” The patient
who does not get to the bathroom or commode
on time or who is incontinent at least once
during the 2-week period of the evaluation is
considered as ‘has occasional accidents”, the
intermediate category. Patients in the third
category are incontinent or controlled by the
supervision, direction, or intervention of another
person. Presence of a catheter or planned super-
vised scheduling for bowel control are included
in the incontinent category.

Feeding concerns the process of getting food
from a plate or its equivalent into the mouth. It
is considered in a primitive sense and without
concern for social niceties. A patient feeds
himself “without assistance” (first of three
classes of feeding on Study Form A-5) when this
primitive process of ingestion is accomplished
without the aid of another person. The

147

 
Effects of Continued Care

intermediate category on Study Form A-5 ap-
plies to the individual who feeds himself, but
receives assistance in cutting meat or buttering
bread. The third category on Study Form A-5,
“receives assistance’, applies to the individual
who is assisted in this feeding process or who is
fed partly or completely parenterally.

In addition to the Index of Independence in
Activities of Daily Living, separate measures of
walking and house-confinement are recorded on
Study Form A-5. Five categories of walking are
identified as follows: (1) “walks on level with-
out assistance’; (2) “walks without assistance
but uses single, straight cane’; (3) “walks with-
out assistance but uses 2 points for mechanical
support, such as crutches, a walker, or 2 canes
(or wears a brace)”; (4) “walks with assistance”;
(5) “uses wheelchair only’; and (6) “neither
walking nor using wheelchair” (Katz, S., Ford,
A.B., Heiple, K.C. and Newill, V.A.: Studies of
Illness in the Aged. Recovery after Fracture of
the Hip. J. Gerontol. 719:285, 1964). These six
categories are defined as mutually exclusive;
thus category “4” applies to patients who walk
with assistance whether they do or do not use
mechanical aids, and category “5” includes only
those who use a wheelchair and do not walk at
all. As indicated before for other functions, the
observer records the most dependent level of
walking which was performed during the 2-week
period of evaluation. A patient who walks about
unassisted in the home is, thus, designated as
“walks with assistance’ if she is assisted or
supervised outdoors. The observer is occasion-
ally told that the patient is not permitted to
walk unless supervised for reasons of safety. In
such cases, direct observation of walking status
is needed to determine whether supervision is
actually present or whether supervision is
claimed merely because assistance is within
hearing distance.

As a measure of patient confinement, observ-
ers record on Study Form A-5 the degree of
house-confinement in the following terms:
(1) “has been outside of residence on 3 or more
days during past 2 weeks”; (2) “has been outside
of residence on only 1 or 2 days during past 2
weeks”; and (3) “has not been outside of resi-
dence in past 2 weeks.” At the time of the initial

148

research evaluation, the period of evaluation is

defined as the 2-week period which preceded the
current period of continuous hospitalization.

The trip home, immediately after intake into the
study, is excluded from this measurement. The
term “outside” means outside of the enclosed,
personal living quarters of the patient. “Out-
side” may be the backyard, terrace, patio, or
porch of the dwelling. It may also be another
apartment or the hall in the building in which
the patient has his apartment suite.

Study Form A-10.

The test of range of movement and strength was
developed to meet the need for an objective,
easily applied, and standardized measure which
could be administered to elderly people. The
test was developed as a simplified form of
previously described, more extended tests
(Daniels, L., Williams, M. and Worthingham, C.:
Muscle Testing Techniques of Manual Examina-
tion, published by W. B. Saunders Company,
2nd Edition, 1961). (Committee on Medical
Rating of Physical Impairment: A Guide to the
Evaluation of Permanent Impairment of the
Extremities and Back. J.A.M.A., Special Edition,
Feb. 15, 1958). Though easily learned by an
examining physician, successful and reliable use
requires careful adherence to the following
definitions and procedures.

The range of a given movement is measured in
degrees and is the angle formed in moving the
limb from a stationary or zero starting point to
the position of completion. In the starting
position, the patient sits erect in a chair with
arms. The chair’s seat is 16 to 18 inches from
the floor, and the supports for the arms of the
chair are recessed at least 4 to 6 inches from the
front corners of the seat. The patient’s head is
positioned high, maintaining the normal curve of
the cervical spine; and the patient sits as erectly
as possible with his back against the back of the
chair. The scapulae are in the normally-adducted
rest position. The upper arms parallel the sides
of the trunk, while the forearms are in a neutral
position, slightly flexed at the elbow and in a
position between pronation and supination. The
hips are flexed with the gluteal region positioned
to the rear of the seat against the back of the
chair. The knees are flexed, and the feet are flat
on the floor with toes pointed directly ahead.

The range of movement for each of the
sixteen maneuvers of the test is defined in
Figure XIII-2 and is expressed as the minimum
range which must be completed in order to
consider that the movement is performed. If the
minimum acceptable range is completed in the
proper positions and directions, the block for
the appropriate maneuver is checked in the
column labeled “Does” on Study Form A-10. If
the patient does not complete the minimum
acceptable range, the block is checked in the
column labeled “Does Not’. In some cases, the
use of a goniometer may be necessary to
measure the angle of movement. For some
patients, also, demonstration of maneuvers is
required.

Fixation is the process by which the appropri-
ate body position is maintained to permit the
muscles involved in a given maneuver to be
tested without the use of other muscles. Ade-
quate fixation ensures that only the muscles of
concern are responsible for the movement and
that substitution of other muscles does not
occur. In Figure XIII-2 the region of fixation or
stabilization is defined for each maneuver. Some
patients cannot maintain the appropriate body
position, and the examiner may have to isolate
the test maneuver by manual stabilization. The
examiner is trained to avoid substitute actions
and is on the alert for even subtle substitutions.
For subjects with adequate strength, certain
maneuvers can be tested in the left and right
limbs spontaneously; and substitution is less
likely to occur.

Resistance is the force exerted by the ex-
aminer in opposition to the patient’s muscle
power. The patient’s power to oppose resistance
is an indicator of patient strength; and, in the
present test, the patient is classed on Study
Form A-10 as performing ‘Against resistance”
when he exerts even slight strength in opposition
to the examiner’s resistance. The examiner’s
force is exerted in a direction opposite to force
vector of the movement being tested, and the
points of application are defined in Figure
XIII-2. Resistance is applied when the patient

Schedule, Forms and Definitions

has completed the minimum acceptable range of
movement.

In this test, the patient is asked to perform a
series of six maneuvers with each lower limb and
10 maneuvers with each upper limb. All maneu-
vers are tested both with and without resistance.
Maneuvers are tested using defined positions for
evaluating movement, defined points for apply-
ing resistance, defined avoidance of substitute
maneuvers, and standardized definitions of range
of movement in degrees. Test maneuvers of the
lower limb include movements of the hip, knee,
and ankle. Patients are instructed as follows:
(1) kick foot into air; (2) bring knee toward
chest; (3) bring leg out to side, (4) return leg to
center; (5) move foot from side to side; and
(6) bend foot up and down. In testing the
function of the upper limb, patients are in-
structed to move the shoulder, elbow, wrist,
hand, and fingers as follows: (1) raise arm
forward high above head; (2) put hand on back
of head; (3) swing arm toward and away from
midline at shoulder level; (4) bring hand to front
of shoulder; (5) with elbow at shoulder level,
straighten out arm; (6) wave hand from side to
side; (7) grasp small object with fingers; (8) ex-
tend fingers; (9) touch tip of thumb with little
finger; and (10) after touching fingertip, extend
thumb. Each movement is demonstrated by the
examiner when necessary.

Observations are recorded on Study Form
A-10 according to the procedures and defini-
tions described herein. For the purposes of the
analyses in this study, function is recorded as
“Does not” whenever the degree of function
defined in Figure XIII-2 is not achieved. The
reason for non-achievement is not considered;
thus, function is recorded as “Does not” for the
patient who does not cooperate, as well as for
the patient with an amputation.

Four grades of function are recognized for
each lower limb; namely, grades L1, L2, L3, and
L4. Nonfunction in at least 5 of the 6 maneuvers
of a lower limb constitutes the lowest grade, L4,
for that limb. Nonfunction in 2 to 4 of the 6
maneuvers of a lower limb is graded as L3. A
grade of L2 is assigned to a lower limb which is
functional in at least 5 of the 6 test maneuvers
and which functions against resistance in no

149
Effects of Continued Care

 

Range of Movement

Region To Be
Stabilized

Point of Application
of Resistance

 

Kick foot into air - knee extended to at least
170° with foot off floor

Thigh proximal to knee

Dorsal area of ankle

 

Bring knee toward chest - thigh proximal to
knee joint lifted at least 10° off seat of chair

Stabilized by trunk

Thigh proximal to knee
joint

 

 

 

 

and downward at least 10° from neutral position

Bring straight leg out to side - leg brought at Pelvis Lateral malleolus

least 15° away from midline with knee as

straight as possible - heel may be dragged on

floor and patient may slouch

Return leg to center - abducted leg moved at Pelvis Medial malleolus

least 15° toward midline with knee as straight

as possible - heel may be dragged on floor and

patient may slouch

Move foot from side to side - dorsiflexed foot Distal tibia Metatarsophalangeal joint of

moved at least 5° from side to side - examiner great toe for inversion and

may support suspended leg off floor MP joint of small toe for
eversion

Bend foot up and down - foot moved upward Distal tibia Dorsum of metatarsals for dor-

siflexion and plantar area of
metatarsals for plantar flexion

 

Raise straight arm forward high above head -
hand and arm raised forward from neutral posi-
tion to a position at least above shoulder level
with elbow extended as much as possible and
with forearm partly or wholly supinated

 

Top of shoulder girdle

Flexor surface of forearm
adjacent to wrist

 

Put hand on back of head - arm raised sidewards
at least 75° with elbow flexed and then hand
raised at least 45° toward occiput - hand need
not touch back of head

Top of shoulder girdle

Dorsal surface of forearm
adjacent to wrist

 

Swing straight arm toward and away from mid-
line at shoulder level - extended arm moved at
least 75° away from and toward midline at
shoulder level with hand partly or wholly supi-
nated - acceptable if extended arm is partly
raised from body though it may not be at
shoulder level

 

Shoulder girdle

Dorsal and flexor surfaces,
respectively, of forearm
adjacent to wrist

 

Bring hand to front of shoulder - elbow flexed
and forearm moved at least 90° toward shoulder
from position in which elbow is extended and
forearm is wholly or partly supinated - examiner
may support upper arm at shoulder level if
necessary

 

 

 

Upper arm adjacent
to elbow

 

Flexor surface of forearm
adjacent to wrist

 

Figure XIII-2. Definitions for Test Range of Movement and Strength

150

 
Schedule, Forms and Definitions

 

With elbow at shoulder level, straighten out arm -

 

forearm extended to within at least 10° of full

Upper arm adjacent
to elbow

Dorsal surface of forearm
adjacent to wrist

extension from position in which elbow is flexed -
examiner may support upper arm at shoulder
level if necessary

 

Wave hand from side to side - wrist flexed ven-
trally at least 25° and dorsally at least 15° with
forearm midway between pronation and supina-
tion - forearm may be supported if necessary

Forearm adjacent to
wrist respectively, at metacarpo-

Palm and dorsum of hand,

phalangeal joints

 

Grasp small object with fingers - object with
circumference of pen grasped and supported

Proximal hand

Each finger pulled lightly

 

Extend fingers - all fingers extended to approxi-
mately 180°

Proximal hand

Dorsal distal tip of each
finger

 

Touch tip of thumb with tip of little finger -
pads of tips of little finger and thumb brought
into contact

 

Proximal hand

Palm at distal end of first
and fifth metacarpals

 

After touching tip of little finger, extend thumb -
thumb extended fully

 

Proximal hand

 

Dorsum of distal phalanx

 

 

 

Figure XIII-2. Continued

more than 1 of the 3 knee and ankle maneuvers.
A grade of L1 is assigned to a lower limb which
is functional in at least 5 of the 6 test maneuvers
and which functions against resistance in at least
2 of the 3 knee and ankle maneuvers.

Three grades are recognized for each upper
limb, namely, Ul, U2, and U3. Nonfunction in
at least 9 of the 10 maneuvers of an upper limb
is graded as U3. A grade of Ul is assigned to an
upper limb, if at least 2 of the 3 shoulder
maneuvers are functional and, at the same time,
if function is present against resistance in at least
6 of the 7 elbow, wrist, fingers, and thumb
maneuvers. A grade of U2 is assigned to an
upper limb which functions better than a grade
of U3 and worse than a grade of Ul.

Study Form A-20

The Highland View Hospital Scale of Psycho-
social Adjustment uses a Q-sort technique to
rate behavior of elderly patients with emphasis
on adjustment to disability and to rehabilitation.
(Shontz, F. C. and Fink, S. L.: A Method for
Evaluating Psychosocial Adjustment of the

Chronically Ill. Am. J. Physical Med. 40. 63-69,
1961). For the measure, each of a series of
thirty-six descriptive statements about behavior
is printed on a single card. The rater, on the
basis of his best knowledge of the patient, sorts
the set of cards into three piles, placing those
statements which best describe the patient to
the right, those which are least like the patient
to the left, and the remaining cards in the
middle. A further breakdown of the cards is
performed until a final distribution is attained in
which seven piles exist with 2,4, 6, 12, 6, 4, and
2 cards in the respective piles from left to right.
The coded identities of the items are recorded in
the corresponding seven columns of Study Form
Number 20. Items which reflect appropriate
adjustment are weighted by factors of 0 to 6 as
specified for the respective columns on Study
Form Number 20, and then summed. Items
which reflect poor adjustment are weighted
similarly and then summed. The difference
between these sums is the total adjustment
score. Subscores can be calculated for subtest
areas of motivation, social adjustment, and
personal adjustment.

151
Effects of Continued Care

Items concerned with motivation are specified
as:

Reasonably cooperative in carrying out
medical procedures,

Comprehends physical condition realisti-
cally,

Moderately active, considering his dis-

ability,

Appraises his capacities accurately,
Thwarts purposes of medical care,
Misinterprets reality of his illness,

Avoids any form of exertion (by desire
rather than necessity),

Thinks he can do anything,

Overly meticulous in obedience to medical

requirements,

Preoccupied with the facts of his illness,

Constantly busy with unproductive activ-

ities,

Devaluates his own abilities.
The first four items in the above list are
considered to be descriptive of appropriate
adjustment.

Items concerned with social adjustment are

specified as:

Family treats him as an equal,
Is accepted by family for what he really is,

Maintains mature social relationships,
Expresses appropriate feelings for family,

Other members of family make all de-
cisions for him,

152

Over-protected at home,

Isolates self from others,

Always fighting with family members,

“Rules the roost” at home,

Clings to others for help,

Family resents his presence,

Overly devoted to his family.
The first four items in the above list are
considered to be descriptive of appropriate
adjustment.

[tems concerned with personal adjustment are
specified as:
Lives as well as possible within the physical
limits of disability,
Emotional reactions are appropriate,
Adaptable in his thinking,
Accepts necessary misfortunes,
Constantly tense,
Apathetic,
Has set ideas he won’t give up,

Blames others whenever something goes
wrong,

Denies disability,
Intellectually chaotic and unpredictable,
“Uses” disability for selfish ends,
Always sees himself at fault.
The first four items in the above list are

considered to be descriptive of appropriate
adjustment.
CHAPTER XIV.

Patients in the study were categorized into
INTAKE CLASSES at the beginning of the
study and into OUTCOME CLASSES during the
course of the study. This served the purposes of
the analyses described in Chapter 12. On the
following pages, short definitions and explana-
tory remarks are presented for 71 intake classes
and 28 outcome classes. The lists of classes were
useful in programming for the study, and they
are reproduced here as aids to the perception of
methods, data, and results of the study. For
additional information about definitions, the
reader should consult the Index of the book.

After every definition in the column labelled
“Class Definition”, two numbers are set off by
parentheses. The first number is the number of
people in the total study group of 300 who were
known to be in the defined class. The second
number is the number of people who were
known to be not in the class. The difference
between the sum of these two numbers and the
total number of patients (300) is the number in
which the class status is not applicable or
unknown. As an example, we consider “intake
class #2: patients with ADL rating of F or G
(92, 208)”. In this example, 92 patients had
ADL ratings of F or G; and 208 patients had

INTAKE AND OUTCOME VARIABLES

other ADL ratings. Since the sum of 92 and 208
accounts for all 300 patients, we deduce that
there were no patients of unknown ADL status
at intake. As another example, we consider
“intake class #61: patients who have daily
contact with no one (59,236)”. In this example,
59 patients were known to have daily contact
with no one; and 236 patients were known to
have daily contact with others. Since the sum of
59 and 236 accounts for 295 patients, we
deduce that the presence or absence of daily
contact was unknown for five patients. For each
outcome class, reasons for excluding patients
from the class are explained in the column
labelled “Remarks” and help to clarify the
meaning of the numbers within the parentheses.
As an example, we consider “outcome class #3:
patients who deteriorated in ADL rating during
the study (36, 176)”. In this example, 36
patients deteriorated in ADL rating during the
study, and 176 maintained or improved their
ADL ratings. Patients who died were excluded
from analysis, and outcome class #2 indicates
that 87 patients died during the study. Since the
sum of 36, 176, and 87 accounts for 299
patients, we deduce that outcome in terms of a
change in ADL rating was unknown for one
patient.

INTAKE VARIABLES

Intake
Class No.
#1 All patients, inclusive
(300, 0)
#2 Patients with ADL rating of F or G
(92, 208)

Class Definition

153
Effects of Continued Care

Intake

Class No.

#3

#4

#5

#6

#7

#8

#9

#10

#11

#12

#13

#14

#15

#16

154

Class definition

Patients not in intake class #2; and with at least one of cancer within 5 years, abnormal
Q-wave on electrocardiogram, or blood urea nitrogen level of 30 mg. per 100 ml. or
greater

(81, 219)

Patients in neither intake class #2 nor #3
(127,173)

Patients not in intake class #2; and with at least one of cancer within 5 years, abnormal
Q-wave on electrocardiogram, blood urea nitrogen level of 30 mg. per 100 ml. or greater,
or on digitalis

(111, 189)

Patients in neither intake class #2 nor #5
(97, 203)

Patients not in intake class #2; and with at least one of cancer within 5 years, abnormal
Q-wave on electrocardiogram, blood urea nitrogen level of 30 mg. per 100 ml. or greater,
or diastolic blood pressure of 90 or greater

(104, 195)

Patients in neither intake class #2 nor #7
(103, 196)

Patients not in intake class #2; and with at least one of cancer within 5 years, abnormal
Q-wave on electrocardiogram, blood urea nitrogen level of 30 mg. per 100 ml. or greater,
on digitalis, or diastolic blood pressure of 90 or greater

(130, 169)

Patients in neither intake class #2 nor #9
(77,222)

Patients not in intake class #2; and with at least one of cancer within 5 years, abnormal

Q-wave on electrocardiogram, blood urea nitrogen level of 30 mg. per 100 ml. or greater,

on digitalis, diastolic blood pressure of 90 or greater, or proteinuria of 2-plus or greater
(133, 166)

Patients in neither intake class #2 nor #11
(74, 225)

Patients with ADL rating of A, B, C, or D
(78,222)

Patients with ADL rating of E, F, or G
(222, 78)

Patients with ADL rating of A, B,C, D, or E
(208, 92)

Patients in intake class #13; and with at least one of Raven test score of 12 or less, or
orientation and mental control test score of 13 or less
(8, 286)
Intake

Class No.

#17

#18

#19

#20

#21

#22

#23

#24

#25

#26

#217

Intake & Outcome Variables

Class definition

Patients in intake class #13; and with neither Raven test score of 12 or less, nor
orientation and mental control test score of 13 or less
(64, 230)

Patients in intake class #14; and with at least one of Raven test score of 12 or less, or
orientation and mental control test score of 13 or less
(67,213)

Patients in intake class #14; and with neither Raven test score of 12 or less, nor orienta-
tion and mental control test score of 13 or less
(135, 145)

Patients in intake class #15; and with at least one of Raven test score of 12 or less, or
orientation and mental control test score of 13 or less
(31, 258)

Patients in intake class #15; and with neither Raven test score of 12 or less, nor
orientation and mental control test score of 13 or less
(166, 123)

Patients in intake class #2; and with at least one of Raven test score of 12 or less, or
orientation and mental control test score of 13 or less
(44, 241)

Patients in intake class #2; and with neither Raven test score of 12 or less, nor orientation
and mental control test score of 13 or less
(33,252)

Patients in intake class #2 with no cancer within 5 years; and with at least one of
abnormal Q-wave on electrocardiogram, blood urea nitrogen level of 30 mg. per 100 ml.
or greater, on digitalis, diastolic blood pressure of 90 or greater, or proteinuria of 2-plus
or greater

(59, 241)

Patients in intake class #15 with no cancer within 5 years; and with at least one of
abnormal Q-wave on electrocardiogram, blood urea nitrogen level of 30 mg. per 100 ml.
or greater, on digitalis, diastolic blood pressure of 90 or greater, or proteinuria of 2-plus
or greater

(115,184)

Patients with at least one of intake class #2, Raven test score of 12 or less, or orientation
and mental control test score of 13 or less; and with no cancer within 5 years; and with at
least one of abnormal Q-wave on electrocardiogram, blood urea nitrogen level of 30 mg.
per 100 ml. or greater, on digitalis, diastolic blood pressure of 90 or greater, or
proteinuria of 2-plus or greater

(77,218)

Patients in intake class #21 with no cancer within 5 years; and with at least one of
abnormal Q-wave on electrocardiogram, blood urea nitrogen level of 30 mg. per 100 ml.
or greater, on digitalis, diastolic blood pressure of 90 or greater, or proteinuria of 2-plus
or greater

(92,202)

155
Effects of Continued Care

Intake
Class No.

#28

#29

#30

#31

#32

#33

#34

#35

#36

156

Class definition

Patients with principal diagnosis of stroke; and with at least one of intake class #2, Raven
test score of 12 or less, orientation and mental control test score of 13 or less, cancer
within 5 years, abnormal Q-wave on electrocardiogram, or blood urea nitrogen level of 30
mg. per 100 ml. or greater

(60, 238)

Patients with principal diagnosis of stroke; and in intake class #21; and with neither
cancer within 5 years, abnormal Q-wave on electrocardiogram, nor blood urea nitrogen
level of 30 mg. per 100 ml. or greater

(15,283)

Patients with principal diagnosis of stroke or chronic brain syndrome; and with at least
one of intake class #2, Raven test score of 12 or less, orientation and mental control test
score of 13 or less, cancer within 5 years, abnormal Q-wave on electrocardiogram, or
blood urea nitrogen level of 30 mg. per 100 ml. or greater

(76,221)

Patients with principal diagnosis of stroke or chronic brain syndrome; and in intake class
#21; and with neither cancer within 5 years, abnormal Q-wave on electrocardiogram, nor
blood urea nitrogen level of 30 mg. per 100 ml. or greater

(18,279)

Patients with principal diagnosis of stroke, chronic brain syndrome, or brain diseases
other than stroke and other than chronic brain syndrome; and with at least one of intake
class #2, Raven test score of 12 or less, orientation and mental control test score of 13 or
less, cancer within 5 years, abnormal Q-wave on electrocardiogram, or blood urea
nitrogen level of 30 mg. per 100 ml. or greater

(90, 207)

Patients with principal diagnosis of stroke, chronic brain syndrome, or brain diseases
other than stroke and other than chronic brain syndrome; and in intake class #21; and
with neither cancer within 5 years, abnormal Q-wave on electrocardiogram, nor blood
urea nitrogen level of 30 mg. per 100 ml. or greater

(21,276)

Patients with principal diagnosis of fracture of femur; and with at least one of intake class
#2, Raven test score of 12 or less, orientation and mental control test score of 13 or less,
cancer within 5 years, abnormal Q-wave on electrocardiogram, or blood urea nitrogen
level of 30 mg. per 100 ml. or greater

(34, 264)
Patients with principal diagnosis of fracture of femur; and in intake class #21; and with
neither cancer within 5 years, abnormal Q-wave on electrocardiogram, nor blood urea

‘nitrogen level of 30 mg. per 100 ml. or greater

(23,275)
Patients with principal diagnosis of fracture of femur or fracture of lower limb other than
fracture of femur; and with at least one of intake class #2, Raven test score of 12 or less,
orientation and mental control test score of 13 or less, cancer within 5 years, abnormal
Q-wave on electrocardiogram, or blood urea nitrogen level of 30 mg. per 100 ml. or

greater
(41,257)
Intake & Outcome Variables

Intake
Class No. Class definition

#37 Patients with principal diagnosis of fracture of femur or fracture of lower limb other than
fracture of femur; and in intake class #21; and with neither cancer within 5 years,
abnormal Q-wave on electrocardiogram, nor blood urea nitrogen level of 30 mg. per 100
ml. or greater

(28, 270)

#38 Patients with principal diagnosis of fracture of femur, fracture of lower limb other than
fracture of femur, or muscular or skeletal conditions other than fracture of femur and
other than other fractures of lower limb; and with at least one of intake class #2, Raven
test score of 12 or less, orientation and mental control test score of 13 or less, cancer
within 5 years, abnormal Q-wave on electrocardiogram, or blood urea nitrogen level of 30
mg. per 100 ml. or greater

(57,239)

#39 Patients with principal diagnosis of fracture of femur, fracture of lower limb other than
fracture of femur, or muscular or skeletal conditions other than fracture of femur and
other than other fractures of lower limb; and in intake class #21; and with neither cancer
within 5 years, abnormal Q-wave on electrocardiogram, nor blood urea nitrogen level of
30 mg. per 100 ml. or greater

(53,243)
#40 Patients with principal diagnosis of arthritis
(31, 269)

#41 Patients with principal diagnosis of cardiovascular-renal diseases; and with at least one of
intake class #2, Raven test score of 12 or less, orientation and mental control test score
of 13 or less, cancer within 5 years, abnormal Q-wave on electrocardiogram, or blood urea
nitrogen level of 30 mg. per 100 ml. or greater

(77,221)

#42 Patients with principal diagnosis of cardiovascular-renal diseases; and in intake class #21;
and with neither cancer within 5 years, abnormal Q-wave on electrocardiogram, nor blood
urea nitrogen level of 30 mg. per 100 ml. or greater

(24,274)

#43 Men
(99, 201)

#44 Women
(201, 99)

#45 Patients younger than 65 years
(75, 225)

#46 Patients 65 through 74 years old
(99, 201)

#47 Patients 75 years old or older
(126, 174)

#48 Patients with Hollingshead Index Social Class of either I, II, or III
(121, 179)

157
Effects of Continued Care

Intake
Class No.

#49

#50

#51

#52

#53

#54

#55

#56

#57

#58

#59

#60

#61

#62

#63

158

Class definition

Patients with Hollingshead Index Social Class of either IV or V
(179, 121)
Patients who own their residence, are employed, and do not receive agency support;
or
Patients who are either unemployed or do not own their residence (not both), and do not
receive agency support
(133,167)

Patients who either receive agency support or do not own their residence (not both), and
are unemployed;
or
Patients who are unemployed, do not own their residence, and receive agency support
(167,133)

Patients who terminated employment or homemaking within the 30 days prior to
admission to Abington House
(112,188)

Patients who terminated employment or homemaking between 30 days and 2 years prior
to admission to Abington House

(98,202)
Patients in neither intake class #52 nor #53

(90, 210)

Patients who are married and live with spouse
(111,189)

Patients who do not live with spouse and do not live alone
(103, 197)

Patients who live alone
(85, 215)

Patients who live with spouse or child
(162, 138)

Patients who have daily contact with two or three categories of persons (spouse, other
relatives, friends)
(71,224)

Patients who have daily contact with one category of persons (spouse, other relatives, or
friends)
(165, 130)

Patients who have daily contact with no one
(59, 236)

Patients who talk to more than 5 people daily
(111, 170)

Patients with any regular organized social activities (church, clubs, etc.)
(134, 164)
Intake
Class No.

#64

#65

#66

#67

#68

#69

#70

#71

Outcome

Class No.

#1

#2

#3

Intake & Outcome Variables

Class definition

Patients who were gainfully employed during the two weeks prior to hospitalization in
Abington House

(45, 255)
Patients in intake class #59, and in at least one of intake class #62, #63, or #64
(50, 245)
Patients in intake class #60, and in at least one of intake class #62, #63, or #64
(105, 191)
Patients in intake class #61, and in at most one of intake class #63 or #64
(42,252)
Patients in intake class #53, or who lost spouse in the 2-year period prior to study
(104, 196)
Patients not in intake class #68
(196, 104)
Patients who talk to fewer persons daily than they did at age 45
(222,71)
Patients not in intake class #70
(71,222)
OUTCOME VARIABLES
Class
Definition Remarks

All patients, inclusive

(300, 0)
Patients who died For interval analyses, this class included all patients, cumula-
during the study tively, who died at any time between intake and the end of each
(87, 213) quarterly interval.
Patients who deterio- Deterioration was any change from left to right in the following
rated in ADL rating graded Index of ADL ratings as defined in Chapter 5: A4, B, C,
during the study D, E, F, G. If the rating was the same or better at terminal than
(36, 176) at intake, there was no deterioration. Excluded from analysis

were those who died and those for whom information was
insufficient. For interval analyses, the comparison was between
ADL ratings at intake and ADL ratings at the end of each
quarterly interval.

159
Effects of Continued Care

Outcome

Class No.

#4

#5

#6

#7

#8

160

Class
Definition

Patients who deterio-
rated in walking
during the study

(30, 182)

Patients who became

more house-confined

during the study
(54,157)

Patients who suffered
a fracture during the
study

(35,193)

Patients who deterio-
rated in upper limb
function during the
study

(27, 185)

Patients who deterio-
rated in lower limb
function during the
study

(26, 185)

Remarks

Deterioration was any change from left to right in the following
graded walking scale as defined in Chapter 5: walking by self or
with no more assistance than a cane, walking with mechanical
assistance, walking with personal assistance (with or without
mechanical assistance), not walking at all. If the walking status
was the same or better at terminal than at intake, there was no
deterioration. Excluded from analysis were those who died and
those for whom information was insufficient. For interval
analyses, the comparison was between walking status at intake
and walking status at the end of each quarterly interval.

Increased house-confinement was any change from left to right
in the following graded classification as defined in Chapter 5:
out of residence 3 or more times during the 2-week reference
period, out 1 or 2 times, out at no time. Excluded from analysis
were those who died and those for whom information was
insufficient. For interval analyses, the comparison was between
house-confinement at intake and house-confinement at the end
of each quarterly interval.

Patients who had fractures during the study, including those
who later died, were counted as being in this class. Those whose
fracture status was unknown or who died without having
suffered a fracture were excluded from analysis. For interval
analyses, this class included all patients, cumulatively, who had
suffered fracture at any time between intake and the end of
each quarterly interval.

Deterioration was any change to a decreased level of function as
measured by the 3 grades of the test of upper limb range of
movement and strength defined in Chapter 5. If the grade was
the same or better at terminal than at intake, there was no
deterioration. Excluded from analysis were those who died and
those for whom information was insufficient. Interval analyses
were not possible, since range of movement and strength were
tested only at intake and at terminal.

Deterioration was any change to a decreased level of function as
measured by the 4 grades of the test of lower limb range of
movement and strength defined in Chapter 5. If the grade was
the same or better at terminal than at intake, there was no
deterioration. Excluded from analysis were those who died and
those for whom information was insufficient. Interval analyses
were not possible, since range of movement and strength were
tested only at intake and at terminal.
Outcome

Class No.

#9

#10

#11

#12

#13

#14

Class
Definition

Patients who deterio-
rated in the function
of any limb during
the study

(43,169)

Patients who deterio-

rated in observation

and clear thinking

during the study
(85,95)

Patients who deterio-
rated in orientation
and mental control
during the study

(68, 128)

Patients who received
financial assistance
from a charitable
agency during the
study, but not at
intake

(13, 198)

Patients who had been

gainfully employed at

intake and were not

employed at terminal
(19,193)

Patients who decreased
their social activities
during the study

(69, 133)

Intake & Outcome Variables

Remarks

Deterioration was any change to a decreased level of function in
any upper or lower limb as measured by the test of range of
movement and strength defined in Chapter 5. If the level of
function was the same or better at terminal than at intake, there
was no deterioration. Excluded from analysis were those who
died and those for whom information was insufficient. Interval
analyses were not possible, since range of movement and
strength were tested only at intake and at terminal.

Deterioration was any change to a lower score according to the
Raven test described in Chapter 5. If the score was the same or
better at terminal than at intake, there was no deterioration.
Excluded from analysis were those who died and those for
whom information was insufficient. Interval analyses were not
possible, since the Raven Test was given only at intake and
terminal.

Deterioration was any change to a lower score according to the
test of orientation and mental control described in Chapter 5. If
the score was the same or better at terminal than at intake,
there was no deterioration. Excluded from analysis were those
who died and those for whom information was insufficient.
Interval analyses were not possible, since the test of orientation
and mental control was given only at intake and at terminal.

Patients who received charitable assistance during the study,
including those who later died, were counted as being in this
class. Those whose assistance status was unknown or who died
without receiving charitable assistance were excluded from
analysis. For interval analyses, this class included all patients,
cumulatively, who had received financial assistance from a
charitable agency through the end of each quarterly interval and
who had not received such assistance at intake as defined in
Chapter 5.

Excluded from analysis were those who died and those for
whom information was insufficient. For interval analyses, the
comparison was between employment status at intake (defined
in Chapter 5) and employment status at the end of each
quarterly interval.

A decrease was any decrease in the number of social activities as
defined in Chapter 5. Excluded from analysis were those who
died and those for whom information was insufficient. For
interval analyses, the comparison was between the number of
social activities at intake and the number at the end of each
6-month interval.

161
Effects of Continued Care

Outcome

Class No.

#15

#16

#17

#18

#19

#20

162

Class
Definition

Patients who decreased

the number of people

to whom they talked

daily during the study
(72, 120)

Patients who decreased

their personal interac-

tion during the study
(60, 139)

Patients who increased
their economic
dependence during
the study

(46, 164)

Patients who were

admitted to nursing

homes during the study
61,177)

Patients who were

hospitalized during

the study
(171,107)

Patients who were

visited by a visiting

nurse during the study
(179, 98)

Remarks

A decrease was any decrease in the number of people to whom
patients talked daily as defined in Chapter 5. Excluded from
analysis were those who died and those for whom information
was insufficient. For interval analyses, the comparison was
between the number of people to whom patients talked at
intake and the number at the end of each 6-month interval.

A decrease was any decrease in the number of categories of
people with whom patients interacted, where the categories
included spouse, other relatives, and friends as defined in
Chapter 5. Excluded from analysis were those who died and
those for whom information was insufficient. For interval
analyses, the comparison was between the number of categories
at intake and the number at the end of each 6-month interval.

An increase was any change to a level of greater dependence as
measured by the four graded classes of the Index of Economic
Dependence defined in Chapter 5. Excluded from analysis were
those who died and those for whom information was insuffi-
cient. For interval analyses, the comparison was between
dependence rating at intake and dependence rating at the end of
each quarterly interval.

Patients who were admitted to nursing homes during the study,
including those who later died, were counted as being in this
class. Those whose nursing home status was unknown or who
died without being admitted to nursing homes were excluded
from analysis. For interval analyses, this class included all
patients, cumulatively, who had been admitted to nursing
homes at any time between intake and the end of each
quarterly interval.

Patients who were hospitalized during the study, including
those who later died, were counted as being in this class. Those
whose hospitalization status was unknown or who died without
being hospitalized were excluded from analysis. For interval
analyses, this class included all patients, cumulatively, who had
been hospitalized at any time between intake and the end of
each quarterly interval.

Patients who were visited by a visiting nurse during the study,
including those who later died, were counted as being in this
class. Those who died before being visited by a visiting nurse
were excluded from analysis. For interval analyses, this class
included all patients, cumulatively, who had been visited by a
visiting nurse at any time between intake and the end of each
quarterly interval.
Outcome

Class No.

#21

#22

#23

#24

#25

#26

#27

#28

Class
Definition

Patients who were
visited by a nurse other
than a visiting nurse
during the study

(38, 255)

Patients who were

cared for in the home

by other than nurses
(173,122)

Patients who were
visited at home by a
physician during the
last 2 weeks of the
study

(22, 189)

Patients who had
physician contacts
during the last 2
weeks of the study
(93,118)

Patients who received
services during the
study from a dentist,
eye doctor, optome-
trist, chiropodist,
social worker, or
physical therapist
(118,87)

Patients who died in a
hospital during the
study

(48, 39)

Patients who were
attended by a physi-
cian on the day of
death

(62, 25)

Fatalities on whom an
autopsy was performed
21, 57)

Intake & Outcome Variables

Remarks

Patients who were visited by a nurse other than a visiting nurse
during the study, including those who later died, were counted
as being in this class. Excluded from analysis were those whose
status with regard to such visits was unknown.

Patients who were cared for by other than nurses during the
study, including those who later died, were counted as being in
this class. Excluded from analysis were those whose status with
regard to such visits was unknown.

Patients who were visited at home by a physician during the last
2 weeks of the study, including those who later died, were
counted as being in this class. Excluded from analysis were
those who died before the 2-week period or whose status with
regard to such visits was unknown. For interval analyses,
analagous counts were obtained for the last 2 weeks of each
quarterly interval.

Patients who had any contact with a physician during the last 2
weeks of the study, including those who later died, were
counted as being in this class. Excluded from analysis were
those who died before the 2-week period or whose status with
regard to such contacts was unknown. For interval analyses,
analagous counts were obtained for the last 2 weeks of each
quarterly interval.

Patients who received any of the indicated professional services
during the last 2 weeks of the study, including those who later
died, were counted as being in this class. Excluded from analysis
were those who died before the 2-week period or whose status
with regard to such services was unknown. For interval analyses,
analagous counts were obtained for the last 2 weeks of each
quarterly interval.

Excluded from analysis were patients who were alive at the end
of the study.

Excluded from analysis were patients who were alive at the end
of the study.

Excluded from analysis were patients who were alive at the end
of the study.

163
APPENDIX:
STATISTICALLY SIGNIFICANT TREATMENT EFFECTS AT END OF STUDY

The significant results at the terminal point of
_ the study are presented on the following pages.
Each line in the list includes, in order, the Intake
Class No. (defined in Chapter 14), the Outcome
Class No. (defined in Chapter 14), and the
numbers of treatment and control patients in
the specific Intake Class according to the pres-
ence or absence of the indicated outcome. From
each line of results, the following model table
can be constructed:

 

Patients in specified Intake Class

Indicated Treatment Control

outcome group group Totals
Present a Cc atc
Absent b d b+d
Totals a+b c+d n

The letters in the model table are defined as
follows:

a = number of patients from the treatment
group who were in the intake class and also in
the outcome class,

b = number of patients from the treatment
group who were in the intake class but not in
the outcome class,

c¢ = number of patients from the control
group who were in the intake class and also in
the outcome class,

d = number of patients from the control
group who were in the intake class but not in
the outcome class, and

 

n = total number of patients.

 

 

Included in the list are results defined as
statistically significant at the 5 per cent level
either by the Chi-square test or by Fishers

164

direct method for calculating probability. Re-
sults which were not significant at the 5 per cent
level by either test were considered nonsignifi-
cant in the text. The criteria used to determine
which test to apply were:

No test was done if either a+b, c+d, atc, or
b+d was less than 4, or if n was less than 13.

Fisher's test was done if n was less than 20, or
if n was less than 40 and the expected value
for any of the four cells was less than 5.

Chi-square test was done in all other instan-
ces.

 

Intake ~~ Outcome
class no. class no a b c¢ d n
1 19 95 46 76 61 278
1 20 145 4 34 94 277
1 28 11 10 2 15 38!
2 19 30 11 20 19 80
2 20 47 0 18 18 83
3 20 39 0 8 27 174
4 4 4 44 13 40 101
4 20 59 4 8 49 120
4 25 36 12 28 22 98
5 9 1 37 7 34 79
5 18 5 35 15 31 86
5 20 53 1 13 35 102
6 4 3 35 10 31 179
6 20 45 3 3 41 92
7 20 48 3 10 35 96
8 7 2 36 9 34 81
8 9 4 34 12 31 81
8 20 SO 1 6 40 97
9 20 62 3 14 41 120
9 23 6 41 1 46 94
10 19 26 10 18 19 73
10 20 36 1 2 34 73
11 20 62 3 14 44 123
11 23 6 41 1 49 97
12 20 36 1 2 31 170
Appendix

Intake Outcome Intake Outcome
classNo. classNo. a bb ¢ dn class No. classNo. a b ¢ d nn
13 9 1 26 10 25 62 34 19 12 4 4 11 31
13 11 6 20 16 16 58 34 20 16 0 3 14 33
13 18 1 26 9 27 63 35 20 8 1 1 12 22
13° 20 . 36 0 5 31 72 36 19 14 7 5 12 38
14 10 14 24 23 12 73! 36 20 21 0 4 15 40
14 19 74 31 50 47 202 37 20 10 1 1 15 27
14 20 109 4 29 63 205 38 8 0 20 5 18 43
14 24 40 35 27 48 150 38 14 9 8 5 18 40
14 25 48 25 34 38 145 38 19 19 9 10 15 53
14 26 25 14 13 19 71 38 20 20 0 6 21 56
14 28 9 8 1 11 29! 39 3 0 16 7 20 43
15 9 8 68 19 63 158 39 4 0 16 7 20 43
15 20 98 4 16 76 194 39 9 1 15 11 16 43
17 7 1 22 7 22 52 39 11 1 15 10 14 40
17 9 1 22 7 22 52 39 20 20 2 2 28 52
17 11 4 18 14 13 49 40 9 2 9 11 6 28
17 18 1 22 7 22 52 40 20 12 1 3 15 31
17 20 30 0 2 27 59 41 20 46 0 11 12 69
18 14 9 9 1 18 37 41 24 19 11 4 16 50
18 20 31 0 13 14 58 41 26 115 2 9 27
19 10 10 18 18 9 55! 42 14 9 2 2 7 20
19 19 47 21 30 29 127 42 20 1m 1 3 7 22
19 20 67 3 14 46 130 43 5 14 20 5 23 62
19 25 33 15 24 25 97 43 6 5 32 0 28 65
20 14 5 6 0 10 21 43 20 51 1 13 22 87
20 17 6 6 0 10 22 43 24 20 14 8 20 62
20 19 11m 4 5 9 29 44 13 6 62 1 81 150
20 20 15 0 4 8 27 44 20 94 3 21 72 190
21 20 78 3 11 66 158 45 7 0 31 6 24 ol
22 20 20 0 11 8 39 45 9 2 29 9 21 61
23 20 19 0 5 7 31 45 11 6 22 13 15 56
24 20 31 0 11 13 55 45 20 34 4 8 25 71
25 20 55 3 12 37 107 46 20 53 0 11 27 O91
26 19 26 10 16 17 69 47 10 10 22 21 14 67
26 20 40 0 15 16 71 47 14 15 20 6 37 78
26 24 14 11 6 17 48 47 16 13 21 7 35 76
27 20 44 2 7 33 86 47 19 39 15 29 30 113
28 20 36 0 10 9 55 47 20 58 0 15 42 115
28 24 15 9 4 13 41 48 20 60 3 5 42 110
29 20 7 0 2 5 14 48 24 24 18 16 29 87
30 20 43 0 13 11 67 49 20 85 1 29 52 167
30 24 15 13 5 15 48 50 5 16 35 6 41 98
31 20 9 0 2 6 17 50 20 70 1 15 38 124
32 3 7 24 1 24 56 50 25 32 18 18 27 95
32 20 49 0 16 15 80 51 14 19 29 12 47 107
32 24 17 14 7 18 56° 51 20 75 3 19 56 153
33 20 110 2 6 19 51 22 33 10 25 21 89!

165
Effects of Continued Care

Intake Outcome Intake Outcome

class No. classNo. a bb ¢ d n classNo. classNo. a b c¢ d n
51 28 11 15 3 19 48 61 28 7 3 0 5 15
52 9 3 32 14 36 85 62 20 57 0 10 38 105
52 13 6 29 2 48 85 63 20 70 3 12 40 125
52 20 48 2 14 40 104 63 26 15 5 5 9 34
53 18 4 33 14 26 77 64 20 25 0 4 14 43
53 20 50 0 13 30 93 65 7 0 23 5 9 37
53 28 7 5 1 10 23 65 20 29 2 6 9 46
54 19 30 15 11 22 78 66 19 36 16 22 23 97
54 20 47 2 7 24 80 66 20 56 1 10 33 100
55 5 18 29 2 28 77 67 2 8 8 5 21 42
55 20 63 3 14 24 104 67 20 15 0 6 17 38
55 24 27 20 10 21 78 68 18 4 33 15 29 81
56 11 5 22 15 20 62 68 19 35 12 27 22 96
56 20 45 1 10 35 91 68 20 51 0 13 34 98
56 25 23 5 20 17 65 69 9 10 56 21 49 136
57 20 37 0 10 34 81 69 20 94 4 21 60 179
58 20 88 3 20 39 150 70 14 25 42 12 65 144
59 5 11 17 2 17 47 70 19 74 28 54 48 204
59 20 40 2 11 13 66 70 20 105 4 27 67 203
59 24 18 10 6 14 48 70 26 27 13 13 20 73
60 17 17 41 7 54 119 70 28 8 7 1 13 29!
60 20 79 2 16 56 153 71 3 8 21 2 27 58
60 25 39 19 25 33 116 71 20 36 0 6 25 67
61 20 24 0 7 24 55

"Included in this line are patients observed only at the interaction effect was present between the observation
beginning and end of the study, since a concomitant process and the treatment program.

166
INDEX

Abington House, an intermediate hospital 16-17, 83 intake classes 98-100, 100, 101, 153-159
background vii-x (Preface), 1-5 interaction effects 84-87
classification 54-57, 68-70, 153-163 description of 80-81, 84-85, 86, 87

(See measures.)

clinical diagnosis 54-55

clinical diagnosis and indicators of disease 56, 56
dementia 57

disability 55-56

disease-disability profiles 56-57
identifying social characteristics 68-69
indicators of disease 55, 57

intake classes 98-99, 153-159
psychological 51-52

outcome classes 102-105, 153, 159-163
social deprivation 69-70

social interaction 69

design of experiment 5-6, 8-14, 4648, 95-105, 107

(See classification.)

(See measures.)

basic plan of 8-10, 95-96

completeness of observations 4748, 48
controlled study 8, 95-96

criteria for inclusion of patients 10, 17

data collection process 13-14, 4647, 107,108
death information, plan for 12

experimental effects 9-10, 96-98

experimental groups 8-9, 96-97

factorial design 9

intake classes, description of 98-99, 153-159
interaction effects 9, 9-10, 96-98

interval observations, plan for 11-12

interview process 46-47

nursing program 8, 9,9-10, 11, 29-31, 96-98
nursing program effects 9, 9-10, 96-98
observation effects 9, 9-10, 96-98

observation schedule /07, 108

organization of staff 12-13

outcome classes, description of 102, 102, 153, 159-163
pilot group 10, 17

randomization, effectiveness of 98-99, 100, 101
randomization process 10-11, 95-96

reliability of outcome data 102, 102

simple intake 8, 10-11

sample selection 16, 17, 83

statistical analysis and interpretation 102-104, 105
terminal observations, plan for 12

treatment program (See nursing program.)
validity of experimental groups 100-102

effects

(See interaction effects.)

(See interval nursing program effects.)

(See observation effects.)

(See terminal nursing program effects.)

(See terminal nursing program effects, specified.)

generalizability 83

design for 9, 9-10, 84, 96-98
interpretation of 84-85
interval nursing program effects 76-82
hospitalization 76, 78-79, 80, 81
house-confinement 79-80
injuries 76-77, 78
mortality 76, 77
nursing home admission 76-78, 79, 81
professional services 79, 81-82
social interaction 79-80
measures 49-52, 109-152
(See classification.)
activities of daily living 50, 55,57, 119-120, 146-148
care 52,121-124,137-144
death 50, 136
disease screening 49-50, 55-56, 56,57,126,128,130-132
Hollingshead Index of Social Class 51, 68, 109-110, 145
house-confinement 50, 720, 148
household composition 51, 69, 116,118,125
Index of ADL 50,1719-120, 146-148
Index of Economic Dependence 51, 68, 715,117,145-146
injury 50, 128,137
intake measures 49-50
nursing activities 140-144
orientation and mental control 51, 735
permanent move /25
physical function 50, 55, 56-57, 119-120,127,129, 146-151,
150-151
principal diagnosis 54-55,56,56,132
psychological 51-52, 133-135, 151-152
psychological adjustment 52, 7133, 151-152
range of movement and strength 50, 7127, 129, 148-151,
150-151
Raven Coloured Progressive Matrices 51, 134
social and economic 50-51, 68-70, 109-118,137, 145-146
social deprivation 51, 69-70,110,112,114,115,117
social interaction 51, 69, 111-114
social role 51, 68-69,109-110,115,117,137, 145
socioeconomic class 51, 68, 109-110,115,117, 145-146
summary of 49
walking 50, 120, 148
methods
(See classification.)
(See design of experiment.)
(See intake classes.)
(See measures.)
(See outcome classes.)
(See sample.)
nursing program in study 28-45
experimental design 8, 9, 9-11, 29-31, 96-98
frequency of visits 31-32, 32, 33-41, 42-44
nursing activities 29, 44-45, 140-144

167
Effects of Continued Care

patients visited 31-32, 3/
sustained visiting 32, 32
visits in illustrative cases 43-44
withdrawal from 42-44, 42
observation effects 83-85
description of 83-84, 85
design for 9, 9-10, 83-84, 96-98
interpretation of 84
outcome classes 102, /02, 153, 159-163
purpose of study 5-6, 16
hypotheses 5-6
results
(See interaction effects.)
(See interval nursing program effects.)
(See observation effects.)
(See terminal nursing program effects.)
(See terminal nursing program effects, specified.)
sample 16-26
description at intake: age 18; sex 18; disease 18, /9; chronic
disease abnormalities 18-20, 20; physical disability
20-21, 21; social and economic 21-25, 22, 24;
psychological 25-26, 26
pilot group 10, 17
selection 16, 17
source, Abington House 16-17

terminal nursing program effects 59-67, 71-75, 87-93, 164-166

(See terminal nursing program effects, specified.)

areas of measurement 59-60

function, injury, and mortality by age 71, 72

function, injury, and mortality for all patients 60

function, injury, and mortality by chronic disease abnormali-
ties 60, 62-64, 64

function, injury, and mortality by diagnosis 60-61, 61, 62

function, injury, and mortality by disease-disability profiles
60, 61-62, 63

168

function, injury, and mortality by identifying social charac-
teristics 71-72

function, injury, and mortality by sex 71

function, injury, and mortality by social deprivation classes
71,72,73

function, injury, and mortality by social interaction classes
71,72,72

infrequent outcomes 66-67

interpretation of 87-93

quantitative estimates 88-91

summarizations 60, 64, 66, 71,73, 75,88-91, 164-166

use of services by age 73, 73

use of services for all patients 64, 65

use of services by chronic disease abnormalities 64, 66-67

use of services by diagnosis 64-66

use of services by disease-disability profiles 64, 66

use of services by identifying social characteristics 73-74

use of services by sex 73-74

use of services by social deprivation classes 73, 74-75, 75

use of services by social interaction classes 73, 74, 74

terminal nursing program effects, specified 60-66, 71-75, 88-91,
164-166

hospitalization 64-66, 65, 73-75, 73-75,89, 91

house-confinement 71-72, 88, 90

limb movement 60-62, 61-64, 71-72, 72-73, 88, 90

nursing home admission 66, 74-75, 89, 91

orientation and mental control 60, 61, 62, 64,71-72,72,
88,90

physician services 65-66, 73-74,89, 91

professional services, nonphysician and non-nurse 66, 74,
89,91

quantitative estimates, summarized 88-91

services at death 64-66, 74-75,89, 91

social interaction 62, 71-73, 88, 90

walking 60,61, 62,63,88, 90

Visiting Nurse Association of Cleveland 28-29

#U.S. GOVERNMENT PRINTING OFFICE: 1973 725-569/343 1-3
 

 
DEPARTMENT OF
HEALTH, EDUCATION, AND WELFARE
PUBLIC HEALTH SERVICE POSTAGE AND FEES PAID
HEALTH SERVICES AND MENTAL HEALTH ADMINISTRATION us. nae OF HEwW
ROCKVILLE, MARYLAND

 

OFFICIAL BUSINESS
PENALTY FOR PRIVATE USE, $300

DHEW Publication No. (HSM) 73-3010
C02930444b