HE SSI 
 
SOME ELEMENTS OF VITAL STATISTICS. 
 
 By Frederick L. Hoffman, LL. D., Statistician, The Prudential 
 Insurance Company of America. 
 
 Within recent years the public health of the nation and the several 
 States has, for the first time in our history, been made the subject of 
 general discussion and has aroused a widespread intelligent public interest. 
 The intimate relation between health and well-being has been precisely 
 established by elaborate investigations, until it may be asserted, without 
 fear of contradiction, that the conservation of health, next to the conser- 
 vation of material resources, is admitted to be of foremost government 
 concern. It requires no extended inquiry to establish the fact that, for an 
 intelligent understanding of many public questions, accurate and 
 uniform vital statistics are indispensable and a prerequisite for intelligent 
 action on the part of the nation, the state or the municipality. Vital- 
 statistics interpreted in this sense, as a matter of national concern, com- 
 prehend all of the essential vital phenomena, that is, births, marriages, 
 deaths, and also the numerous but non-fatal diseases which afflict man- 
 kind and curtail the normal duration of human life. 
 
 From early times the importance of registering the essential vital facts 
 of human experience has been recognized by law givers or law-making 
 bodies and the registration of vital statistics in New England, at least, is 
 almost coincident with the time when the settlement of the country had 
 assumed the status of permanency in community life. Obviously the 
 registration of births, marriages and deaths is a governmental function 
 which cannot be delegated to non-governmental agencies, although in 
 Catholic countries this function is often properly performed as an essen- 
 tial element in Church administration. Passing over the earlier efforts in 
 the New England States, which have more historical rather than practical 
 interest, it maj' be stated that the first effective American law on the sub- 
 ject was passed by the Massachusetts Legislature on the third of March, 
 1842, the clerks of the several towns and cities in the Commonwealth 
 being required to transmit to the Secretary of State a certified copy of 
 th'eir records of the births, marriages and deaths of all persons within 
 their respective towns and cities. 
 
 The first annual report of the vital statistics of Massachusetts was pub- 
 lished in 1843, but it was not until fifteen years later that the first corre- 
 sponding report for the State of Vermont was published, containing the 
 returns of births, marriages and deaths, for the year ending on the thirty- 
 first day of December, 1857. The first registrar was Dr. Hiram S. 
 Stevens of St. Albans, who had taken a deep interest in the subject, and to 
 whose influence the State is indebted for the passage of the act requiring 
 
 c 
 
 1 
 
the returns. The first report on the vital statistics of Vermont was from 
 the pen of Dr. Chas. L. Allen of Middletown. The report was prepared 
 in conformity to a plan recommended by the American Medical Associa- 
 tion, which, from its origin, has taken a most active and intelligent interest 
 in advancing the cause of vital statistics as a prerequisite for the ascer- 
 tainment of the underlying causes and conditions affecting the public 
 health. 
 
 Vital statistics have their legal, medical, social, economic and commercial 
 aspects, all of which embrace matters of serious concern to the public at 
 large. The legal identification of the person frequently involves important 
 pecuniary considerations, as well as serious questions of public policy. 
 The medical aspects of vital statistics are essential in advancing the cause 
 of medical science. Unless medicine, as a healing art, is intelligently co- 
 ordinated to the geographical distribution of disease, it must fail, at 
 least to some extent, in achieving the best possible results. The social 
 and economic condition of a given community is reflected in its vital 
 statistics, and in the fluctuations in its birth, marriage, death and morbidity 
 rates, many of which can be coordinated to corresponding fluctuations in 
 the economic and moral condition of the population. Equally so, the 
 material progress of the nation, as conditioned by the industrial efficiency 
 of wage earners, is more or less a question of health and longevity, and 
 by means of precise calculations it is now possible to at least approxi- 
 mately estimate the economic loss resulting from the occurrence of pre- 
 ventable disease, the reduction or elimination of which is recognized to- 
 day as a matter of community concern. The conservation of wage 
 earners' health and strength is a factor in our industrial progress, the 
 importance of which has lately been brought out in a memorial address 
 to the President of the United States by a special committee of the 
 American Association for Labor Legislation, suggesting the appointment 
 of a national commission to ascertain the facts and recommend a course 
 of action in conformity to our needs. 
 
 The foregoing brief statement of the legal, medical and economic 
 importance of vital statistics emphasizes the duty of all whose services are 
 ' required to establish a system of registration, at once trustworthy in its 
 - details, complete in its scope, and clearly interpreted by means of qualified 
 statistical and medical analysis. Primarily the duty of accurate registra- 
 tion rests upon the physician, as the one public functionary whose intimate 
 knowledge of the surrounding circumstances has from the outset imposed 
 upon him the public duty to make record of facts. The importance of 
 some of these facts may be difficult to recognize, as, for example, the 
 accurate registration of births, but they may assume most far-reaching 
 consequences in course of time. It is true that the filling out of birth and 
 death certificates often imposes an arduous task upon the practitioner, 
 already over-burdened with the most serious responsibilities which can 
 fall to the lot of anyone. But the duty of preserving the vital experience 
 
 
of a commonwealth transcends all others in importance, when every fact 
 and condition which has a bearing upon the question is intelligently taken 
 into account. While, therefore, the duty cannot be shifted and requires 
 to be discharged with absolute fidelity to the high purpose for which the 
 system of vital registration has been established, it is clearly the corre- 
 sponding duty of the community to make reasonable compensation for 
 such services, but in the nature of the case, the pecuniary compensation 
 can never be entirely adequate to the labor involved. 
 
 Next to entire completeness of the vital records of births and deaths, 
 there is the duty of the attending physician to fill out, to the fullest extent 
 of his knowledge, the certificate of death, so that all of the medical facts 
 which have a bearing upon the termination of a human life may be a 
 matter of complete and trustworthy record for all time. The importance 
 of accurate medical diagnosis of the causes of death is not limited to the 
 individual, nor to the time being, but, for comparative purposes, the dis- 
 tribution of deaths from principal causes at one period may throw im- 
 portant light upon the corresponding distribution of deaths from the same 
 causes at another period, and enable those who make a study of the sub- 
 ject to ascertain the underlying factors, without a knowledge of which a 
 diminution in the preventable mortality is practically out of the question. 
 
 Since every problem of public health is partly a question of comparison 
 in local conditions, it is of the utmost importance that there should be a 
 reasonable degree of uniformity, both in the use of standard certificates 
 of death and standard methods of death classification. The certificate 
 adopted by the Division of Vital Statistics of the Census Office is, for this 
 purpose, the most suitable, since it has the approval of the American 
 Public Health Association, aside from the fact that it is the basis of the 
 tabulation and analysis of the mortality of about 60 per cent of the total 
 population of the Uhited States. Uniformity of classification and the tabu- 
 lar analysis of the deaths in conformity to the Bertillon system are essen- 
 tial, in that this classification has international sanction and is used by most 
 of the governments of the civilized countries of the earth. If, therefore, a 
 state desires to make the largest possible practical use of its vital statistics, 
 it is necessary that the use of a standard certificate of death be insisted 
 upon, and it is equally important that the Bertillon classification of causes 
 of death be used, as that classification has been accepted from the outset 
 by the Division of Vital Statistics of the Census Office, and is now in 
 use by many of the American States and by most of the more important 
 cities. 
 
 The analysis of vital statistics is a most delicate and difficult task, which, 
 unfortunately, has often of necessity to be performed by those ill qualified 
 for so important a public duty. The calculation of birth, marriage and 
 death rates requires to be made in conformity to well-established prin- 
 ciples which are often disregarded by those in charge of the registration, 
 tabulation and analysis of the vital statistics of the different states and 
 
municipalities. The publication of the results rarely conforms to stan- 
 dardized methods and there is, apparently, no definite tendency towards 
 a decided improvement, except in so far as order has been brought out 
 of chaos by the publication of the national vital statistics for the registra- 
 tion area by the Division of Vital Statistics of the Census Office. It would 
 be advisable, however, for American States and cities to adopt, as far as 
 practicable, the standard method of tabulation and analysis sanctioned by 
 the Census Office. This would also be in conformity to the method 
 adopted by the Local Government Board of England, which requires the 
 use of certain standard tables in the annual reports of the local medical 
 officers of health. But perhaps the most serious fault which impairs the 
 utility of our vital statistics is the crude method of calculating relative or 
 proportionate rates of fecundity, mortality and morbidity, and it is only 
 too common to meet with the expression "percentage per thousand," which, 
 of course, is a self-evident and absurd contradiction. It would seem best 
 to calculate all mortality rates on the basis of 1,000 population and the 
 mortality from specific causes on the basis of 10,000 population. Infantile 
 mortality rates should be calculated on the basis of 1,000 births and 
 marriage rates should be calculated on the basis of the number of un- 
 married men and women and separately for each sex. Fecundity rates 
 should be calculated on the number of women fifteen years of age and 
 over and the specific intensity of mortality should be determined by divi- 
 sional periods of life. For a country like the United States, where the 
 elements of the population vary widely, it is absolutely essential, for 
 purposes of accuracy, to reduce the use of crude birth and death rates to 
 a minimum and to calculate such rates with a due regard to age, sex, race, 
 nativity, occupation, etc. Corrected death rates are advisable and not 
 difficult to calculate, and their use is practically essential in states where 
 the population distribution diverges as much from the normal as, for 
 illustration, is true of Vermont. In corrected death rates the age and 
 sex distribution of the population of diff^erent communities is equalized 
 and, for illustration, while the crude death rate of London, Eng., in 1910, 
 was 12.71, the corrected death rate was 13.6. More important differ- 
 ences are brought out by correction as, for illustration, in the case of 
 Oldham, Eng., a typical industrial community, where the crude death rate 
 was 17.25 and the corrected death rate 19.44. The social and economic 
 significance of the birth rate has only been recognized in the United States 
 within comparatively recent years, and for comparatively few states and 
 cities are trustworthy returns available. That there has been a material 
 decline in the birth rate is a fact which hardly requires to be sustained by 
 statistical evidence, but the social and even political significance of this 
 decline becomes apparent when the decrease in fecundity is shown sepa- 
 rately for the native and the foreign-born elements. Some very interest- 
 ing facts have been brought to light by the Immigration Commission, the 
 researches of which show that the birth rate has materially declined, but 
 
much more so for the native-born element than for the foreign-born, 
 and that the average number of children to a family is least for the 
 native-born of native stock. The most important investigations into the 
 decline of the birth rate have been made in the State of New South 
 Wales, by a Royal Commission, the results and conclusions of which 
 challenge the attention of the civilized world. 
 
 Birth rates include a consideration of legitimacy and illegitimacy, of 
 multiple births (twins, triplets, etc.), and the rather difficult problem 
 of still births, for which an exact definition would be a much-desired 
 improvement which would greatly increase the accuracy of comparative 
 vital statistics. It has become the almost universal custom to exclude 
 still births from the calculation of both birth and death rates, and it is 
 of importance that states and cities not following this method should 
 introduce a change, so that the respective rates may be comparable with 
 those of other communities and states. 
 
 The true birth rate of the United States is not accurately known and all 
 estimates are partly a matter of conjecture. Efforts are being made to 
 improve our national registration of vital statistics in this respect, and the 
 Census Office is entitled to the heartiest cooperation on the part of phy- 
 sicians and health officers throughout the land. Assuming that the birth 
 rate of the United States is only 30 per 1,000 per annum, the number of 
 births is approximately 2,760,000 for 1910, and of this number probably 
 not less than 2 per cent are illegitimate. The ratio of plural births is 
 about 2 per cent of the total births, of which perhaps one tenth represents 
 other than twin born. The approximate percentage of still births is 
 probably from three to five, but accurate information is not available. 
 The proportion of male still births is almost invariably greater than the 
 corresponding proportion of female still births. 
 
 I can only very briefly consider the marriage rate, which is usually cal- 
 culated on the basis of .every 1,000 of the total population. Since every 
 marriage involves two persons, the true marriage rate is one half of the 
 number of persons married per 1,000 of population, and calculated by this 
 method, the approximate marriage rate for the United States is probably 
 between 10 and 12.5 per 1,000. The relation of marriages to marriageable 
 population is calculated with some difficulty, since the facts are only 
 available for census years, but approximately the rates are from 60 to 75 
 per 1,000 for males and from 45 to 60 per 1,000 for females. Another 
 method is to calculate the mean age at marriage, which requires, however, 
 the taking into consideration of the sex of the parties to the marriage and 
 whether the marriage was the first or subsequent to the first. For Massa- 
 chusetts in 1909 the average age of all bridegrooms was 29.04 years and 
 of all brides 25.83 years; of men marrying for the first time 27.34 years 
 and of women marrying for the first time 24.60 years. The material 
 differences disclosed by this comparison are of sufficient importance to 
 warrant the calculation of the average age at marriage in the manner 
 
suggested. In Massachusetts the average age at marriage has always 
 been rather high but there has been no decided change in the last thirty 
 years. Corresponding information for the State of Vermont is not 
 available. For the United States as a whole it may be estimated that the 
 marriage rate is about 20 persons per 1,000 of population per annum 
 and on this basis the number of persons marrying during 1910 was 
 approximately 2,144,000. * 
 
 The death rate of a community will always constitute the most impor- 
 tant index factor of physical health and sanitary well-being. As yet it has 
 not been found feasible anywhere to collect complete morbidity statistics, 
 which are not only essential for a full understanding of the problems of 
 health and longevity, but which throw light upon many problems, the solu- 
 tion of which is hopeless upon the basis of mortality statistics alone. The 
 death rate measures the relative incidence of mortality in a given popula- 
 tion and it is almost invariably expressed in the proportion of the number 
 of deaths occurring among a thousand population. The older method of 
 expressing the relative mortality in the form of a ratio of the number of 
 inhabitants to one death is now obsolete and rarely used. Since the death 
 rate varies with every year of life, it is self-evident that the age distribu- 
 tion of the population is of the utmost importance in the calculation of 
 crude death rates and that, therefore, extreme caution is necessary in 
 comparing the mortahty rate of communities which may, or may not, be 
 entirely unlike in their age and sex distribution. Since the mortality 
 rate is highest at the two extremes of life, that is, at ages under five 
 and at ages over sixty-five, it is obvious that a population containing 
 a large proportion of the young, or of the aged, or both, as the case 
 may be, may have a high crude death rate without being in any sense 
 an unhealthy community, or subject to abnormal local ill-health-producing 
 conditions. The difficulty is best overcome by calculating death rates 
 for divisional periods of life, that is, for ages under five, five to four- 
 teen, fifteen to forty-four, forty-five to sixty-four, and sixty-five and 
 over. This method of calculation, of course, requires a knowledge of 
 the age distribution of the population, as derived from the decennial 
 census returns. While the calculation of specific death rates by divi- 
 sional periods of life is not difficult it is rare that the method is 
 used in this country, although it is quite generally used in English cities, 
 which have attained to a most enviable position in matters which pertain 
 to an intelligent local health administration. Death rates, to be trust- 
 worthy, require an accurate return of the population and a complete 
 return of all the deaths occurring in the given community, but, unfortu- 
 nately, there is a tendency to exaggerate, to overestimate, the population 
 for other than census years, and often there is a neglect to secure a com- 
 plete return of all the deaths, including those of non-residents. The object 
 
 ♦The number of divorces is about 88,000 per annum or about 176,000 persons per 
 annum. 
 
of all statistical investigations into matters of public health is not pri- 
 marily to prove a community healthy, or otherwise, but simply to ascertain 
 the truth, as an essential basis for intelligent and effective health admin- 
 istration. 
 
 The causes of death require to be accurately and fully stated on the death 
 certificate, and the analysis needs to be made with painstaking care to 
 avoid often totally erroneous conclusions. In the case of deaths compli- 
 cated by two or even more contributory causes it is best to adopt the 
 Budapest system for purposes of tabular presentation, while the practice 
 of the Census Office should be followed in giving the proper preference 
 of one cause over another, where the system of single classification is 
 followed. The advantages of the Budapest system are that it permits of 
 a full understanding of all the elements of mortality as conditioned by the 
 causes, so that, for illustration, it is shown how many deaths from kidney 
 diseases are complicated by alcoholism, and how many deaths from preg- 
 nancy are complicated by tuberculosis. For medical as well as general 
 purposes, the Budapest system is incomparably superior to the present 
 practice of stating the mortality by single causes only. 
 
 Deaths from violence may here be referred to as a branch of vital 
 statistics which is peculiarly a matter of public concern. Accident, homi- 
 cide, suicide and legal executions form this group, which in the aggregate 
 accounts for about 7.6 per cent of the deaths from all causes. In 1910 
 it is estimated that there were 81,490 deaths from accident, 5,253 deaths 
 from homicide, including executions, and 15,462 deaths from suicide. Of 
 the accidents probably 35,000 were industrial accidents, including about 
 5,000 deaths of railway employees and 3,000 deaths of miners. The accu- 
 rate registration of homicide and suicide involves peculiar difficulties on 
 account of the fact that it is sometimes impossible to ascertain whether 
 death was caused by an accident, by murder, or by self-murder, as the 
 case may be. This is particularly the case in drowning fatalities and to 
 some extent also with gunshot wounds and poisoning. However, when 
 due care is used and where the coroner system is thoroughly efficient, the 
 element of error can be reduced to an almost negligible minimum. Homi- 
 cides are much more frequent in this country than in most of the other 
 civilized countries of the world, and as regards lynching, the United 
 States is almost in a class by itself. Suicides are gradually increasing and 
 the rate for one hundred American cities has changed from 12.3 per 10,000 
 of population in 1890 to 19.7 in 1910. 
 
 The distribution of mortality varies widely with different countries 
 and localities and according to periods of time. From a medical point of 
 view, it is of no small practical importance to know the relative frequency 
 of fatal, as well as non-fatal, diseases, and, of course, the same is true 
 for purposes of public health administration. To illustrate this point, 
 it may be stated that during the decade ending with 1909 the principal 
 cause of death in the State of Vermont was nervous diseases, account- 
 
ing for 13.3 per cent of the total mortalitj', while the principal cause 
 of death in the cit}- of New York and for the same period of time was 
 respirator}- diseases, accounting for 17.6 per cent of the deaths from all 
 causes. The second most important cause in the State of Vermont was 
 respiratory diseases, accoimting for 13.2 per cent, and the third was cir- 
 culatory diseases, accoimting for 13.1 per cent. These three groups of 
 causes, therefore, that is, nervous, respirator}- and circulator}- diseases, 
 accotmt for 39.6 per cent of the deaths from all causes in the State of 
 Vermont. In the citj" of Xew York the second most important cause was 
 tubercular diseases, and the third, digestive diseases, the three groups of 
 causes accounting for 43.8 per cent of the deaths from all causes. The 
 relative distribution of deaths from principal causes for the State of 
 Vermont and the city of New York is shown in the accompan}-ing diagram, 
 which is self-explanatory. (See plate, page 9.) 
 
 It is a matter of opinion as to the proportion of the total mortality 
 which is due to strictly preventable causes. Leaving out of present con- 
 sideration the acute infectious diseases, it is generally admitted that the 
 two diseases of most importance from the standpoint of public health are, 
 first, tuberculosis, and second, typhoid fever. In the State of Vermont 
 tuberculosis, during the decade ending with 1909, caused 8.5 per cent of 
 the deaths from all causes, while t}-phoid fever caused 1.4 per cent, or 
 the two causes combined account for not quite 10 per cent of the total. 
 The relative importance of typhoid fever and tuberculosis is but inade- 
 quately exhibited in the mortality rates, for the economic cost of sickness 
 for both of these diseases is very large. For typhoid fever we have fairly 
 trustworthy statistics, warranting the conclusion that by present methods 
 of treatment from 10 to 20 per cent of the cases will end in death. For 
 tuberculosis our information is quite unsatisfactory and as yet compul- 
 sorv' notification has failed to produce statistics that can be accepted as 
 entirely trustworthy. The average duration of a fatal case of tuber- 
 culosis, however, is probably not less than two years, so that the expense 
 of sickness is of enormous economic importance, considering the fact 
 that there are not less than 150,000 deaths from this disease in the United 
 States every year. The local incidence of tuberculosis and typhoid in the 
 State of Vermont is shown in the diagram, on which the several counties 
 have been arranged in order of the death rate prevailing during the five 
 years ending with 1909. For tuberculosis the death rate was highest in 
 the county of Washington, where it attained to 170.7 per 100,000 of popu- 
 lation, and lowest in the county of Essex, where it was only 92.1. The 
 average for the state was 130.1. (See plate^ page 10.) 
 
 The typhoid fever death rate was highest in Caledonia county, where it 
 attained to 33.8 per 100,000, and it was lowest in Lamoille county, where it 
 was only 7.9. The average rate for the State was 17.2. 
 
 In the case of tuberculosis, there are so many factors which complicate 
 the local incidence of the disease, that even the most painstaking inquiry 
 
 8 
 
Ill 
 
 
 lllllllli 
 
 III 
 
 oo>lvtDir>m';tco — coco 
 
 COCMCVJOJ'-'- — — — '-'-'- 
 
 - XM 01 
 
 o < u. 
 
 en 
 
 •— <0 ■— .= 
 
 O (T u ^ o ^ ^ 
 
 TJ O 
 
 .E E 
 
 
 
 O 
 
 
 in LO o O 
 
 '^f 00 CO CO 
 
 OQ U 
 
 
 CD 
 
 
 •♦- 
 
 
 <0 
 
 a> 
 
 -4— 
 
 ion 
 
 (/) 
 
 c 
 
 
 <o 
 
 <l> 
 
 1- 
 
 
 O 
 
 H 
 
 J2 § -^ ^ 
 
 
 a ^ 
 
 o 
 
 CO <M r-« i« ^ O' r^ 
 OJ OJ — — — — O 
 
 O 
 
 
 < UJ 
 
into the causes of its prevalence will often fail to suggest the most prac- 
 tical means or methods of prevention. Aside, however, from the recog- 
 nized element of direct infection of one person by another, due to personal 
 contact inevitable under unregulated methods of home treatment, the one 
 factor which always stands out as the most suggestive within the scope 
 of public health administration, is dust in any one of its many forms, but 
 particularly industrial dust, as met with in the so-called dusty trades. 
 
 It is most significant that the county of Washington, where the tuber- 
 culosis death rate is 170.7 per 100,000 and the highest in the State of 
 Vermont, should also be the center of the stone-cutting industry, than 
 which there is hardly another occupation more dangerous to human health 
 and life. The continuous and considerable inhalation of stone dust is a 
 direct contributory cause, although the dust itself does not convey the 
 germ of tuberculosis to the lungs. The dust irritates the bronchial tubes 
 and the lungs to such an extent that the lodgment of the bacilli of tuber- 
 culosis takes place under conditions most favorable for its rapid spread 
 and most difficult for its eradication. It requires no extended observation 
 of the conditions under which the stone industry is carried on, particularly 
 since the introduction of pneumatic tools, to bring home the conviction that 
 t|lis industry must be decidedly injurious to health and that its effective 
 supervision and control is one of the foremost considerations of a suc- 
 cessful public health administration in localities where this industry is 
 carried on to any considerable extent. The proof of this assertion is 
 found in every investigation which has been -made into the mortality of 
 persons employed in the different branches of the stone industry and it 
 has been shown that at ages twenty-five to thirty-four, for illustration, 
 from 40 to 60 per cent of the deaths among men in these employments are 
 caused by tuberculosis. Considering the importance of this industry to 
 the State of Vermont, it would seem that \he mortality among stone 
 workers should receive special consideration, in much the same manner as 
 this has been done for the stone industry of Scotland in the annual re- 
 ports of the health officer of Aberdeen and for textile workers in the 
 annual reports of the health officer of Blackburn. It is a matter of regret 
 that the statistics of mortality by occupation published during the earlier 
 years of registration, and also for the five years ending with 1905, should 
 be of such limited utility for the purpose of a qualified inquiry into the 
 occurrence of tuberculosis among the stone workers of Vermont. 
 
 It would carry me too far to discuss in detail other important elements 
 of mortality and I can only briefly refer to the variations in the death rate 
 by age and sex, which are best illustrated by the vital statistics of Eng- 
 land and Wales. There are few exceptions to the rule that the mortality 
 of males is higher than the corresponding mortality of females. At all 
 ages, for illustration, the death rate of males for England and Wales in 
 1909 was 15.4 per 1,000, against 13.7 for females. The rates differ by 
 divisional periods of life, but in England and Wales* the death rates of 
 
 11 
 
females are below those of males at every age. This, however, is not 
 true for a number of other countries, where the mortality of females 
 during the early child-bearing period is above that of males. In Hungary, 
 for illustration, at ages fifteen to nineteen the death rate of males is 5.98 
 per 1,000, against a rate of 1 ^2) for females, and at ages twenty to twenty- 
 four the respective rates are 8.55 and 9.42. 
 
 The variations in the death rate by divisional periods of life according 
 to the returns for England and Wales for 1909, but for males only, have 
 been as follows: Commencing with a rate of 40.3 per 1,000 for the age 
 period under five, the rate was 3.2 for ages five to nine, and only 1.9 at 
 ages ten to fourteen ; subsequent to this age the rate increased to 2.8 for 
 the period fifteen to nineteen, to 3.7 during twenty to twenty-four, to 5.3 
 during twenty-five to thirty-four, and to 9.1 during thirty-five to forty- 
 four. After this age the rise is more rapid : to 16.7 per 1,000 at ages 
 forty-five to fifty-four; to 32.8 at fifty-five to sixty-four; to 73.5 at sixty- 
 five to seventy-four, to 141.8 at seventy-five to eighty-four, and to 321.4 
 per 1,000 at ages eighty-five and upwards. 
 
 I direct attention to these variations to emphasize the utility of calculat- 
 ing specific death rates as the only conclusive method of determining the 
 true incidence of mortality in any given locality or state, and furthermore 
 the necessity of taking into account the age and sex distribution of the 
 population in the calculation of general death rates. It is often said that 
 a death rate of ten per 1,000 would be equivalent to every person living 
 100 years, but this would not necessarily be true and certainly not for a 
 community with a transitory population of an abnormal age distribution, 
 resulting, for illustration, from the presence of a large educational insti- 
 tution, a military school, or a soldiers' home. The first two would ac- 
 count for a considerable proportion of the population at an age period 
 when the death rate is only two to four per 1,000, while the latter would 
 represent a class among whom the death rate would be 15.0 per 1,000, or 
 more. A stationary population, however, is practically never met with in 
 actual experience, and certainly not in the United States, except perhaps 
 in the case of very isolated mountainous communities, for which no sta- 
 tistical information is available. 
 
 There is, no doubt, a definite relation between the birth rate and the 
 death rate, which, however, requires a very careful consideration of the 
 elements of the population, the proportion of child-bearing women, the 
 percentage of still births, plural births and illegitimate births, aside from 
 methods of infant feeding, etc. It is only within recent years that the 
 problem of infant mortality has attracted general attention in this country, 
 but the efforts which are being made to reduce the death rate at young 
 ages are certain to be followed by far-reaching results. Intelligent action 
 depends upon a thorough understanding of the underlying facts which 
 govern the mortality of infants, but enough is known to warrant the con- 
 
 *Same in the United States for 1900. 
 
 12 
 
elusion that a high mortality is largely the result of improper methods of 
 feeding and of a more or less contaminated milk supply. 
 
 The infantile mortality is usually determined by ascertaining the pro- 
 portion of deaths of children under one year to every thousand births 
 during the same year, but in the absence of accurate birth statistics, it is 
 better to calculate the deaths in proportion to the living population under 
 one, or in the form of a percentage of the deaths from all causes. The 
 wide variation in the infantile death rate is emphasized in the compara- 
 tive statistics of certain foreign countries and, for illustration, for the five 
 years ending with 1905 the rate was 215 for Austria and 212 for Hungary, 
 against 138 for England and Wales, 114 for Ontario, and only 75 for New 
 Zealand. In 1909 the rate for England and Wales was only 109, for 
 Ontario 92, and for New Zealand 62. 
 
 Next to infant mortality in importance rank acute infectious diseases of 
 early childhood, which account for a considerable proportion of the mor- 
 tality from all causes. For diseases of this class we have a considerable 
 amount of additional information concerning morbidity, and the most 
 useful statistics of this kind are to be found in the annual reports of the 
 Superintendent of Health of the city of Providence. According to the 
 Providence reports, the fatality in diphtheria cases, for illustration, during 
 the period 1884-1909 was 15.77 per. cent, but comparing the first year of the 
 period with the last, the percentage has been reduced from 30.5 to 8.29. 
 The highest fatality ratio occurred in 1888, when it reached 42.17 per cent. 
 
 Considering the fatality ratios in diphtheria by periods of life it is 
 shown by the Providence returns that the ratio was highest at ages one to 
 two, when it reached 40.27 per cent of all reported cases against 38.79 at 
 ages under one. The maximum number of deaths, however, occurred at 
 ages two to five, while the maximum number of cases occurred at ages 
 two to ten. While the fatality ratio at ages two to five was 25.7 per cent, 
 it was 12.1 per cent at ages five to ten. The ratio declined to 4.9 per cent 
 at ages ten to nineteen, and 2.8 per cent at ages twenty and over. These 
 averages are for the period 1889-1908. The Providence returns contain 
 much additional information with reference to primary and secondary 
 cases, and of the attack rate by ages and the mortality by sex, but these 
 cannot be considered in detail on this occasion. 
 
 In scarlet fever the fatality ratio, according to the Providence returns, 
 was 7.73 per cent for the period 1884-1909. During the first year of the 
 period the ratio was 10.59 per cent, and during the last 1.60 per cent. The 
 ratio was highest in 1888, when it reached 21.88 per cent, and lowest in 
 1889, when it was only 1.24 per cent. Considered by ages, the fatality ratio 
 was highest at ages one to two, having been 25.84 per cent, against 18.18 
 per cent at ages under one. The maximum number of deaths occurred 
 at ages two to five and the maximum number of cases at ages five to ten. 
 The fatality ratio at ages two to five was 12.24 per cent and at ages five 
 to ten, 5.62 per cent. For adults the fatality ratio was 5.7 per cent. 
 
 13 
 
A consideration of the acute infectious diseases of childhood leads to 
 the important question of effective medical inspection of schools. The 
 subject is so vast and complex that it cannot be dealt vv^ith on this occa- 
 sion. Some progress has been made in this country in bringing about uni- 
 form methods of examination, but it is doubtful whether the returns for 
 different localities are strictly comparable. The field of medical inspection 
 is rather ill-defined and much more is done in some localities than in 
 others. The field should not be limited to specific infectious diseases, but 
 made to include diseases of the oral and respiratory tracts, the ears, the 
 eyes, the skin, and miscellaneous ailments, physical defects and deformi- 
 ties. No such inspection can be considered complete unless it compre- 
 hends systematic physical measurements of children, at least as regards 
 height and weight, with a due regard to age. The best reports of medical 
 inspection of school children are made by English medical officers of 
 health for practically all the larger cities of England, where due consid- 
 eration is given to the vast importance of accurate anthropometric ob- 
 servations in determining the physical progress or deterioration of the 
 race. We have only made a crude beginning in this country and the 
 material which has been collected has only been published in a frag- 
 mentary form, but the practical difficulties are greater with us on account 
 of the many different races and nationalities. The difficulties, however, 
 only emphasize the importance of another problem which lies at the root 
 of our national progress in health and well-being. 
 
 In conclusion, I may refer briefly to the importance of life tables for the 
 several states and the larger cities, the construction of which, however, 
 in the United States is a much more involved task, on account of our 
 more heterogeneous and more mobile population as compared with the 
 populations of European countries and states, for which such tables have 
 been calculated from time to time, for many years. Efforts are being made 
 to calculate a life table for the registration area of the United States, 
 which will furnish the first trustworthy basis of comparison with other 
 countries for which such tables are available. For a state like Vermont it 
 would seem less difficult to calculate such a table .than for states with a 
 larger immigrant population, but a fact which complicates the situation 
 is the migration of a relatively large proportion of Vermont-born persons 
 to other states. A notable contribution to the study of the population 
 statistics of Vermont is a historical and statistical study by William S. 
 Rossiter, recently published by the American Statistical Association. 
 According to that study, while the population of Vermont in 1900 was 
 343,641, the total number of natives of Vermont throughout the United 
 States was 416,672, so that 40.4 per cent of Vermont-born persons reside in 
 other parts of the United States than in Vermont. Considerations like 
 these emphasize the difficulty of constructing trustworth}' life tables, and 
 thej^ apply also to the calculation of trustworthy death rates by divisional 
 periods of life. 
 
 14 
 
It had been my intention to discuss somewhat more in detail the 
 broadening field of medical science and the solution of social and economic 
 problems. There has been a remarkable development of public interest 
 in problems of preventive medicine and many of the arguments advanced 
 in behalf of the expenditure of large sums of money are based upon eco- 
 nomic grounds and the claim that in the long run such expenditures will 
 bring a manifold return. The economic cost of tuberculosis and typhoid 
 fever is a matter of almost mathematical demonstration, but it requires 
 no argument to prove that the state or community sustains a definite 
 economic loss through the untimely deaths of wage-earners and others 
 usefully employed in productive industry. Vast expenditures in behalf of 
 a campaign for the reduction of the mortality from tuberculosis, or the 
 elimination of sources of typhoid fever, or for the purification of the 
 public milk supply in behalf of a campaign against infant mortality, 
 require, however, most careful supervision and control, to avoid waste of 
 money and effort upon methods and results which are likely to be based 
 upon superficial reasoning. The question is not so much how public 
 interest can be aroused in matters of sanitary reform, as to how an 
 interest once once aroused can be actively maintained as an indispensable 
 factor in efforts to secure the largest amount of public good. An intelli- 
 gent public opinion is a most valuable asset in all matters of sanitary 
 reform, but perhaps in no direction is this aid as useful as in the fields of 
 industrial hygiene, the medical inspection of- factories and the medical 
 inspection of schools. There is also the additional aid to be derived from 
 the further development of insurance medicine, which has done much and 
 can do much more to advance the cause of all that pertains to the con- 
 servation of public health and the preservation of human life. Every 
 effort of this kind, however, must of necessity rest upon a trustworthy 
 basis of vital statistics, and for this reason the qualified tabulation and 
 analysis of the vital records of the state is a duty of the utmost impor- 
 tance, which may well enlist the efforts of all interested in the intelligent 
 development of the conditions under which we live. 
 
 The following is an abstract of the mortality of Vermont for the fifty 
 years ending with 1909. The record is divided into two periods of 
 twenty-five years each, the aggregate population of the first period having 
 been 8,195,373 and for the second period 8,528,416. The aggregate popula- 
 tion represents the sum of the estimated populations for each year of the 
 fifty-year period. During the first twenty-five years there were 123,386 
 recorded deaths and during the second 143,413. The average death rate for 
 the first twenty-five years was 14.9 per 1,000 and for the second 16.8. The 
 increase is due largely to more complete registration and partly to an 
 increasing proportion of persons sixty-five years of age and over. In 
 Vermont this proportion was 5.5 per cent in 1900, against 4.1 per cent for 
 the United States, and 2.8 per cent for the city of New York. The com- 
 parison is limited to twelve important causes affecting the mortality of 
 
 15 
 
II III nil 
 
 013 736 635 6 
 
 adult life, which constituted 44.3 per cent of the mortality from all causes 
 during the first twenty-five years, and 47.5 per cent during the last quarter 
 century. The three causes, the mortality from which decreased, are 
 typhoid fever, tubercular diseases and diseases of parturition, while the 
 nine causes which show an increased mortality, partly because of an 
 increasing proportion of persons sixty-five years of age and over, are 
 cancer, diabetes, circulatory diseases, pneumonia, liver diseases, genito- 
 urinary diseases, accidents, homicides and suicides. 
 
 A Balance Sheet of Mortality' 
 
 State of Vermont 
 
 1860-1909 
 
 Causes 
 
 J 860-1 884 
 
 1885-1909 1 
 
 Number 
 
 Per 
 
 Per cent. 100 000 
 of All Popu- 
 Causes lation 
 
 Number 
 
 Per ceni 
 of All 
 Causes 
 
 Per 
 . 100,000 
 Popu- 
 lation 
 
 Typhoid Fever 
 
 5,555 
 
 4.5 67.8 
 
 2,345 
 
 1.6 
 
 27.5— 
 
 Tubercular Diseases 
 
 20,887 
 
 17.1 254.9 
 
 14,26(5 
 
 99 
 
 167.3— 
 
 Cancer 
 
 3,152 
 
 2.6 38.5 
 
 6,046 
 
 42 
 
 70.9+ 
 
 Diabetes 
 
 502 
 
 0.4 6.1 
 
 1,062 
 
 0.7 
 
 12 5 + 
 
 Circulatory Diseases 
 
 6,150 
 
 5.0 75.0 
 
 14,8;8 
 
 10.4 
 
 174.6+ 
 
 Pneumonia 
 
 9,015 
 
 7.4 110.0 
 
 13,970 
 
 9.7 
 
 163.8+ 
 
 Liver Diseases 
 
 1,220 
 
 1.0 14.9 
 
 1,668 
 
 1.2 
 
 19.6+ 
 
 Genito-Urinary Diseases 
 
 2,074 
 
 1.7 25.3 
 
 7,064 
 
 4.9 
 
 82.8+ 
 
 Parturition 
 
 1,227 
 
 1.0 14 9 
 
 1,196 
 
 0.8 
 
 14.0— 
 
 Accidents 
 
 3,671 
 
 3.0 44.8 
 
 5,030 
 
 3.5 
 
 59.0+ 
 
 Homicides 
 
 52 
 
 0.1 0.6 
 
 65 
 
 0.1 
 
 0.8+ 
 
 Suicides 
 
 618 
 
 0.5 7.5 
 
 870 
 
 0.6 
 
 10.2+ 
 
 Population— 1860-1884— 8 
 
 .195,373 
 
 4- Increase 
 
 in death rate, 1885-1909 
 
 1885-1909—8 
 
 ,528,416 
 
 — Decrease 
 
 in death rate, 1885-1909 
 
 16 
 
LIBRARY OF CONGRESS 
 
 HoUingo: Corp. 
 pH8.5