key: cord-0861461-r9sb8a06
authors: Greenwald, S. D.; Chamoun, N. G.; Manberg, P. J.; Gray, J.; Clain, D.; Maheshwari, K.; Sessler, D. I.
title: Covid-19 and Excess Mortality in Medicare Beneficiaries
date: 2021-04-10
journal: nan
DOI: 10.1101/2021.04.07.21254793
sha: eb48aac272353add69a861efdce7fa9e5c84c0bf
doc_id: 861461
cord_uid: r9sb8a06

We estimated excess mortality in Medicare recipients with probable and confirmed Covid-19 infections in the general community and amongst residents of long-term care (LTC) facilities. We considered 28,389,098 Medicare and dual-eligible recipients from one year before February 29, 2020 through September 30, 2020, with mortality followed through November 30th, 2020. Probable and confirmed Covid-19 diagnoses, presumably mostly symptomatic, were determined from ICD-10 codes. We developed a Risk Stratification Index (RSI) mortality model which was applied prospectively to establish baseline mortality risk. Excess deaths attributable to Covid-19 were estimated by comparing actual-to-expected deaths based on historical comparisons and in closely matched cohorts with and without Covid-19. 677,100 (2.4%) beneficiaries had confirmed Covid-19 and 2,917,604 (10.3%) had probable Covid-19. 472,329 confirmed cases were community living and 204,771 were in LTC. Mortality following a probable or confirmed diagnosis in the community increased from an expected incidence of about 4% to actual incidence of 7.5%. In long-term care facilities, the corresponding increase was from 20.3% to 24.6%. The absolute increase was therefore similar at 3-4% in the community and in LTC residents. The percentage increase was far greater in the community (89%) than among patients in chronic care facilities (21%) who had higher baseline risk. The LTC population without probable or confirmed Covid-19 diagnoses experienced 38,932 excess deaths (35%) compared to historical estimates. Limitations in access to Covid-19 testing and disease under-reporting in LTC patients probably were important factors, although social isolation and disruption in usual care presumably also contributed. Remarkably, there were 31,360 fewer deaths than expected in community dwellers without probable or confirmed Covid-19 diagnoses, representing a 6% reduction. Disruptions to the healthcare system and avoided medical care were thus apparently offset by other factors, representing overall benefit. The Covid-19 pandemic had marked effects on mortality, but the effects were highly context-dependent.

The Covid-19 pandemic has profoundly influenced US healthcare, especially among 47 considered people residing in the community from those in long term care facilities, who 105 are expected to have a higher baseline mortality risk and thus may be especially 106 susceptible to Methods 108 Data analysis was conducted on the Center for Medicare and Medicaid Services (CMS) 109

Research Identifiable File (RIF) data using SAS Enterprise Guide (Version 7.15) under 110 a special Data Use Agreement (DUA). This project was determined to be exempt from 111 informed consent requirements by the New England Institutional Review Board. Final 112 data analysis of the full cohort was conducted from January 10 to March 11, 2021. This 113 study followed the Strengthening the Reporting of Observational Studies in 114 Epidemiology (STROBE) reporting guideline for cohort studies. (12) 115 Individual subject data used for our analysis are available to certain stakeholders as 116 allowed by federal regulations and CMS policy. Requests for access to data to replicate 117 these findings require an approved research protocol and DUA with CMS. For more 118 information, contact the Research Data Assistance Center (ResDAC, 119 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. ; https://doi.org/10.1101 https://doi.org/10. /2021 We used full Medicare fee-for-service and dual eligible (Medicaid and Medicare) files 126 one year before the anchor date through September 30, 2020, with mortality outcomes 127 reported through November 30 th , 2020 for the primary analysis of We identified beneficiaries with confirmed Covid-19 diagnoses consistent with CMS 129 guidance using ICD-10-CM codes for Covid-19 (B97.29 before April 1, 2020 and U07.1 130 thereafter) as a primary or secondary diagnosis between March 1, 2020 and September 131 30, 2020.(13) Probable Covid-19 infection cases were identified using ICD-10-CM 132 codes consistent with the CDC guidance (Z20.828) and WHO recommendations 133 (U07.2). (14, 15) Presumably most subjects with Covid-19 diagnoses were symptomatic, 134

although some may have been tested because of risk or exposure. applied to all eligible Medicare or dual-eligible beneficiaries as of February 29, 2020 to 159 derive individual RSI scores as of that date -that is, before Covid-19 infections were 160 confirmed in the United States. 161 For comparative purposes, a second model was similarly developed to predict nine-162 month mortality from the presence of 27 individual chronic conditions as defined by 163 CMS.(17) Specifically, logistic regression (stepwise selection using p-in of 10 -3 , p-out of 164 9

We conducted a progression of complementary inquiries. To test our primary 171 hypothesis, we first identified the main study cohorts of beneficiaries with diagnoses of 172 probable or confirmed Covid-19, and then subdivided them based on location of service 173 (community or LTC/SNF) as of February 29, 2020. Within each cohort, we determined 174 9-month mortality between the anchor date and November 30, 2020. 175 Our goal was to first define associations between baseline demographic characteristics 176 and health status as characterized by RSI with the risk of mortality following a Covid-19 177 diagnosis in the overall at-risk population and in pre-defined subpopulations. We initially 178 compared differences in mean baseline RSI scores between survivors and non-179 survivors, then used univariable and multivariable regression modeling to estimate the 180 relative importance of baseline demographic factors, chronic conditions, and RSI scores 181 as independent predictors of mortality. A similar analysis was conducted to identify risk 182 factors associated with a confirmed diagnosis of Covid-19. We also determined the 183 association between RSI and observed mortality by beneficiary age group and location. 184

Two independent methods were used to estimate expected 2020 mortality in our study 186 population. A historical comparison allowed us to compare year-over-year changes in 187 mortality in Medicare recipients and thus characterize overall effects of Covid-19 and 188 quarantine-induced restrictions in healthcare access on mortality. A case-matched 189 analysis provided an alternate estimate of Covid-19-related excess mortality within the 190 9 months of 2020 that we considered. 191 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 10, 2021. 

A second method used case matching or "digital twinning" to estimate excess mortality 207 in exposed subjects compared to concurrent controls who had closely matched health 208 profiles. Beneficiaries receiving a diagnosis of probable or confirmed Covid-19 were 209 pairwise exactly matched 1:1 on Feb 29, 2020 with beneficiaries without a Covid-19 210 diagnosis based on sex, age (within 1-year), ethnicity, location of services in Feb 2020 211 (community or LTC/SNF), along with RSI as a propensity matching factor (within 0.1%). 212 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 10, 2021. were associated with increasing mortality in a consistent rank ordered manner across 251 each age group, thereby suggesting that RSI provides a significant and sensitive 252 measure of co-morbidities and mortality risk that is independent of age. 253 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. ; is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. Confirmed Covid-19 cases were identified consistent with CMS guidance using ICD-10-CM codes for Covid-19 (B97.29 258 before April 1, 2020 and U07.1 thereafter) as a primary or secondary diagnosis between March 1, 2020 and September 259 30, 2020. 12 Probable Covid-19 infection cases were identified using ICD-10-CM codes consistent with the CDC guidance 260 (Z20.828) and WHO recommendations (U07.2). 13,14 The baseline risk of 9-month mortality defined by the Risk 261 Stratification Index (RSI) calculated on February 29, 2020 was 3.2% in the entire population and significantly higher 262 among those who died compared to those who survived. 263 .

It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. Subjects were categorized as "LTC/SNF" if they received services in either a Long-Term Care (LTC) or Skilled Nursing 268 Facility (SNF) in February 2020, otherwise they were categorized as receiving services in the "Community." Confirmed 269

Covid-19 cases were identified consistent with CMS guidance using ICD-10-CM codes for Covid-19 (B97.29 before April 270 1, 2020 and U07.1 thereafter) as a primary or secondary diagnosis between March 1, 2020 and September 30, 2020. 12 271 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. (Table  286 3). 287 is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. ; empty parenthesis under the adjusted odds ratio. RSI, age, and location of services 300 were the strongest (unadjusted) predictors of mortality. RSI and age remain strong 301 predictors following adjustment; however, risks associated with having chronic 302 conditions were typically reduced when adjusted by the presence of RSI and other 303 factors. Status of Lung cancer and end-stage renal disease appear to carry meaningful 304 incremental risk after adjustment. 305 306 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. ; is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. Subjects were categorized as "LTC/SNF" if they received services in either a Long-Term Care (LTC) or Skilled Nursing 309 Facility (SNF) in February 2020, otherwise they were categorized as receiving services in the "Community." Confirmed 310 .

It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. Covid-19 cases were identified consistent with CMS guidance using ICD-10-CM codes for Covid-19 (B97.29 before April 311 1, 2020 and U07.1 thereafter) as a primary or secondary diagnosis between March 1, 2020 and September 30, 2020. 12 312

Probable Covid-19 infection cases were identified using ICD-10-CM codes consistent with the CDC guidance (Z20.828) 313 and WHO recommendations (U07.2). 13,14 The baseline risk of 9-month mortality defined by the Risk Stratification Index 314 (RSI) calculated on February 29, 2020. Beneficiaries receiving a diagnosis of probable or confirmed Covid-19 were 315 pairwise exactly matched 1:1 on Feb 29, 2020 with beneficiaries without a Covid-19 diagnosis based on sex, age (within 316

1-year), ethnicity, location of services in Feb 2020 (community or LTC/SNF), along with RSI as a propensity factor (within 317 0.1%). The tabulated results demonstrate similarity of baseline characteristics between tightly matched populations. The 318 baseline risk of mortality (RSI) was much higher in patients categorized as LTC/SNF than Community subjects. 319 320 321 .

It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021.

The distribution of observed and expected mortality by diagnosis, category, and location 323 of care is presented in Fig 2. As expected, subjects with high baseline mortality risk in 324 the LTC/SNF cohort had actual mortality that exceeded all other groups. Those with 325 confirmed Covid-19 showed similarly increased mortality above expected levels in both 326 the LTC/SNF and community setting. Among community dwelling subjects, mortality 327 also exceeded expected risk in subjects with possible Covid-19. 328

Confirmed Covid -19 cohorts in community, LTC/SNF, and combined analysis. (case matching method) deaths attributable to probable or confirmed Covid-19 across 344 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. ;  the full population in the 9 months of 2020 that we considered. In the matched analysis, 345 half the deaths (50,793) occurred in patients with a confirmed diagnosis of 689) occurred in those with a probable Covid-19 diagnosis. In contrast, 31,360 347 fewer subjects without a Covid-19 diagnosis died than expected, representing a 6% 348 mortality reduction ( Table  4) . 349 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. ; is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. Subjects were categorized as "LTC/SNF" if they received services in either a Long-Term Care (LTC) or Skilled Nursing 357 Facility (SNF) in February 2020, otherwise they were categorized as receiving services in the "Community." Confirmed 358

Covid-19 cases were identified consistent with CMS guidance using ICD-10-CM codes for Covid-19 (B97.29 before April 359 1, 2020 and U07.1 thereafter) as a primary or secondary diagnosis between March 1, 2020 and September 30, 2020. 12 360

Probable Covid-19 infection cases were identified using ICD-10-CM codes consistent with the CDC guidance (Z20.828) 361 and WHO recommendations (U07.2). 13,14 The baseline risk of 9-month mortality defined by the Risk Stratification Index 362 (RSI) calculated on February 29, 2020. Two independent methods were used to estimate expected 2020 mortality as 363 described in the footnote above. The case matching (digital twin) method utilized the baseline risk of 9-month mortality 364 defined by the Risk Stratification Index (RSI). In this method, beneficiaries receiving a diagnosis of probable or confirmed 365

Covid-19 were pairwise exactly matched 1:1 on Feb 29, 2020 with beneficiaries without a Covid-19 diagnosis based on 366 sex, age (within 1-year), ethnicity, location of services in Feb 2020 (community or LTC/SNF), along with RSI as a 367 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. cases were community living and 204,771 were in LTC. Mortality following a probable or confirmed diagnosis in the 371 community increased from an expected incidence of about 4% to actual incidence of 7.5%. In long-term care facilities, the 372 corresponding increase was from 20.3% to 24.6%. The absolute increase was therefore similar at 3-4% in the community 373 and in long-term care residents. But the percentage increase was far greater in the community (89%) than among patients 374 in chronic care facilities (21%) who had high baseline risk. The long-term care population without probable or confirmed 375

Covid-19 diagnoses experienced 38,932 excess deaths (35%) compared to historical estimates. Limitations in access to 376

Covid-19 testing and disease under-reporting in long-term care patients probably contributed, although social isolation 377 and disruption in usual care presumably also contributed. Remarkably, there were 31,360 fewer deaths than expected in 378 community dwellers without probable or confirmed Covid-19 diagnoses, representing about a 6% reduction. Disruptions to 379 the healthcare system and avoided medical care were thus apparently offset by other factors, representing overall benefit. 380

The Covid-19 pandemic had marked effects on mortality, but the effects were highly context-dependent. 381 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021.

Our model is highly predictive for mortality in Medicare beneficiaries with documented 383

Covid-19 infections. Because baseline RSI scores can help to identify Medicare 384 beneficiaries at highest risk for mortality due to Covid-19, we make the models publicly 385 available in the following formats: 386 387 1) Access to RSI risk calculators will be provided free of charge for authorized 388 non-commercial uses via the HDAI API website (https://www.hda-389 institute.com/api/). Age, sex, care location, and comorbidities were significant predictors of mortality. The 401 strongest individual predictor following a diagnosis of Covid-19 across all age 402 categories, and in both community and long-term care settings was the integrated 403 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. ; measure of patient co-morbidities, RSI. Although many individual chronic conditions 404 were also significant risk factors in our unadjusted univariable analysis, the strength of 405 Overall, our historical model indicated that mortality following a probable or confirmed 421 diagnosis in the community increased from an expected incidence of about 4% to actual 422 incidence of 7.5%. In LTC/SNF's, the corresponding increase was from 20.3% to 423 24.6%. Therefore, the absolute increase in mortality was similar at 3-4% in the 424 community and in long-term care residents. However, baseline risk (RSI) associated 425 with all individuals in a care setting varied greatly, being only about 2.6% in the 426 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. ;  community versus 20.5% in long-term care facilities. As a percentage, the relative 427 increase in mortality was thus far greater in the community (89%) than among patients 428 in long term care facilities (21.0%). 429

Somewhat remarkably, overall mortality decreased in Medicare participants without 430 probable or confirmed Covid-19 diagnoses. In fact, among community dwellers, there 431 were 31,360 fewer deaths than expected, representing about a 6% reduction. 432

Disruptions to the healthcare system and avoided medical care were thus apparently 433 offset by other factors, representing overall benefit. Obvious health benefits of pandemic 434 isolation include reduced exposure to other airborne illnesses such as influenza, fewer 435 driving accidents and fewer homicides. However, none seems sufficient to explain the 436 reduction. More subtle effects including reduced work or stress-related illness might 437 contribute more, although there is no obvious reason to believe that the pandemic 438 would reduce stress -especially in an over-65-year-old population. 439

The causes of reduced mortality in community dwelling Medicare participants remains 440 unclear. However, our results suggest that inadequate care for chronic conditions and 441 delayed care of acute events did not produce the feared outcome of higher short-term 442 mortality in the general population without Covid-19. But due to limited follow-up, we 443 caution that disruptions in healthcare delivery may yet result in adverse longer-term 444 outcomes due to delays in the diagnosis and treatment of new and existing chronic 445

conditions. An additional consideration is that prolonged sequela after severe Covid-19 446 infections (Long Covid syndrome) appear substantial and is an area requiring urgent 447 further study. (18) 448 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. ;

There was a distinct disparity between community dwellers and those in long-term care 449 facilities with respect to historical mortality comparisons. In contrast to community 450

Medicare participants, the long-term care population without probable or confirmed 451

Covid-19 diagnoses experienced 38,932 excess deaths (35%) compared to historical 452 estimates. We believe that limitations in access to reporting in long-term care patients probably were responsible for this finding. It seems 454 likely that many of the excess deaths in this vulnerable population were consequent to 455 undiagnosed Covid-19 infections. But it is also probable that social isolation and 456 disruption in usual care may have contributed as well. The higher-than-expected level of 457 excess deaths observed in this cohort (subjects without a probable or confirmed Covid 458 diagnosis) is reflected in our case matching results, which indicate a modest relative 459 reduction in deaths in subjects with a Covid related diagnosis. This is most likely due to 460 undiagnosed Covid cases included in the control population, but we cannot rule out the 461 possibility that the focus on care for the Covid patients had an unintended adverse 462 impact on the remaining population. 463 464

We excluded less than 2.2% of the available population because of missing and 466 inconsistent values. Because data were missing non-systematically, exclusion of these 467 subjects was unlikely to introduce meaningful bias. We relied on administrative 468 diagnostic claims for Covid-19 to assign exposure. Surely these are inexact, especially 469 during our study period early in the pandemic. Furthermore, a new diagnostic code for 470 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. Mortality following a probable or confirmed Covid-19 diagnosis in the community 491 increased from an expected incidence of about 4% to actual incidence of 7.5%. In long-492 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity.

The copyright holder for this preprint this version posted April 10, 2021. ; term care facilities, the corresponding increase was from 20.3% to 24.6%. The absolute 493 increase was therefore similar at 3-4% in the community and in long-term care 494 residents. But the percentage increase was far greater in the community (89%) than 495 among patients in chronic care facilities (21%) who had high baseline risk. 496

The long-term care population without probable or confirmed Covid-19 diagnoses 497 experienced 38,932 excess deaths (35%) compared to historical estimates. Limitations 498 in access to Covid-19 testing and disease under-reporting in long-term care patients 499 probably contributed, although social isolation and disruption in usual care presumably 500 contributed. Remarkably, there were 31,360 fewer deaths than expected in community 501 dwellers without probable or confirmed Covid-19 diagnoses, representing about a 6% 502 reduction. Disruptions to the healthcare system and avoided medical care were thus 503 apparently offset by other factors, representing overall benefit. 504

The Covid-19 pandemic had marked effects on mortality, but the effects were highly 505 context-dependent. Among community dwelling Medicare participants with suspected or 506 confirmed Covid-19 diagnoses, mortality nearly doubled, but from a relatively low 507 baseline. Patients in long-term care facilities had a similar absolute increase in mortality, 508 but because their baseline mortality was 20.5%, the relative increase was smaller. In 509 contrast, community dwelling Medicare participants without COVID had about 6% lower-510 than-expected mortality. 511 512 513 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 10, 2021. . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 10, 2021. Wood Johnson Foundation provides support for providing access to the Covid-19 RSI 365 537 risk prediction models by academic researchers for non-commercial purposes via the 538 HDAI API (https://www.hda-institute.com/api/). 539

Role of Funder/Sponsor: Other than the authors and Data Analyst Douceur Tengu, no 540 additional members of Health Data Analytics had a role in the design and conduct of the 541 study; collection, management, analysis, and interpretation of the data; preparation, 542 review, or approval of the manuscript; or decision to submit the manuscript for 543 publication. 544

We thank John Parks, Zhenyu Hong and Douceur Tengu (Data Analysts, Health Data 546 Analytics Institute) for their assistance in developing models and preparing figures and 547 tables. 548 . CC-BY-NC 4.0 International license It is made available under a is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review)

The copyright holder for this preprint this version posted April 10, 2021. ; https://doi.org/10.1101/2021.04.07.21254793 doi: medRxiv preprint

Excess Mortality in California During the Coronavirus Disease 2019 Pandemic

Excess Mortality and COVID-19-Related Mortality Among US Adults Aged 25-44 Years

Excess Deaths Associated with COVID-19

Medical Conditions and Risk of for Severe COVID-19 Illness: Centers for 563 Disease Control and Prevention

Comorbidities 570 associated with mortality in 31,461 adults with COVID-19 in the United States: A 571 federated electronic medical record analysis

Analyses of Risk, Racial Disparity, and Outcomes 573

Among US Patients With Cancer and COVID-19 Infection

Prediction models for diagnosis and prognosis of covid-19 infection: systematic review 577 and critical appraisal

Model for Severe Covid-19 in the Medicare Population: A Tool for Prioritizing Scarce 580

Vaccine Supply

Study of 534,023 Medicare Beneficiaries with COVID-19: Implications for Personalized 583

Risk Prediction

CDC. ICD-10-CM Official Coding and Reporting Guidelines

Centers for Disease Control and Prevention; 2020. 593 15. WHO. COVID-19 coding in ICD-10: World Health Organization

Health Organization

Validation and Calibration 597 of the Risk Stratification Index

Condition Categories -Chronic Conditions Data Warehouse: Centers for 599

More than 50 Long-term effects of Covid-19: a systematic review and 603 meta-analysis. medRxiv preprint

Uptake 605 and Accuracy of the Diagnosis Code for COVID-19 Among US Hospitalizations

Outcomes 608 patients. The PPV, sensitivity and relative risk (RR) are tabulated for these detector 659 operating points

Comparison of RSI (Panel A) and Chronic Condition based models

No Covid-19, Probable Covid-664 19 and Confirmed Covid-19 populations. Confirmed Covid-19 cases were identified 665 consistent with CMS guidance using ICD-10-CM codes for Covid-19 (B97

Probable Covid-19 infection cases were identified 668 using ICD-10-CM codes consistent with the CDC guidance (Z20.828) and WHO 669 recommendations (U07.2). 13,14 (A,B) ROCs display the sensitivity vs. 1 -specificity in 670 detecting patients who died within 9 months after prediction from February 29,2020 671 (baseline). The areas under each ROC, with their corresponding 95% confidence 672 intervals, are tabulated in the lower right of each figure. Predictions using RSI yielded 673 better performance (A) than those using a model based on age

12 Subjects were categorized as "LTC/SNF" if they received 682 services in either a Long Term Care (LTC) or Skilled Nursing Facility (SNF) in February 683 2020, otherwise they were categorized as receiving services in the "Community

Predictors were assessed at baseline (February 29, 2020) and include quintiles of Risk 685

Stratification Index (RSI), presence of chronic conditions, location of services

age, sex, race, and quintiles of median 687 guidance (Z20.828) and WHO recommendations (U07.2). 13,14 Subjects were 705 categorized as "LTC/SNF" if they received services in either a Long-Term Care (LTC) or

Skilled Nursing Facility (SNF) in February 2020, otherwise they were categorized as 707 receiving services in the "Community

Covid diagnosis, mortality rates were lower in the community setting compared to those 710 in the LTC/SNF; however, for subjects with confirmed or probable Covid infection, 711 mortality rates were typically higher in the community setting than in the