key: cord-0992954-j509mdxo
authors: Chang, Man-Huei; Moonesinghe, Ramal; Truman, Benedict I.
title: COVID-19 Hospitalization by Race and Ethnicity: Association with Chronic Conditions Among Medicare Beneficiaries, January 1–September 30, 2020
date: 2021-01-08
journal: J Racial Ethn Health Disparities
DOI: 10.1007/s40615-020-00960-y
sha: a7d36c47c0daae302a0f27c3bbcf46bfc389d05d
doc_id: 992954
cord_uid: j509mdxo

OBJECTIVES: We assessed the association between hospitalization for illness from COVID-19 infection and chronic conditions among Medicare beneficiaries (MBs) with fee-for-service (FFS) claims by race and ethnicity for January 1–September 30, 2020. METHODS: We used 2020 monthly Medicare data from January 1–September 30, 2020, reported to the Centers for Medicare and Medicaid Services to compute hospitalization rates per 100 COVID-19 MBs with FFS claims who were hospitalized (ICD-10-CM codes: B97.29 before April 1, 2020; ICD-10-CM codes: U07.1 from April 1, 2020, onward) with or without selected chronic conditions. We used logistic regression to estimate adjusted odds ratios with 95% confidence intervals for association of person-level rate of being hospitalized with COVID-19 and each of 27 chronic conditions by race/ethnicity, controlling for age, sex, and urban-rural residence among MBs. RESULTS: COVID-19-related hospitalizations were associated with all selected chronic conditions, except osteoporosis and Alzheimer disease/dementia among COVID-19 MBs. The top five conditions with the highest odds for hospitalization among COVID-19 MBs were end-stage renal disease (adjusted odds ratios (aOR): 2.15; 95% CI: 2.10–2.21), chronic kidney disease (aOR: 1.54; 95% CI: 1.52–1.56), acute myocardial infarction (aOR: 1.45; 95% CI: 1.39–1.53), heart failure (aOR: 1.43; 95% CI: 1.41–1.44), and diabetes (aOR: 1.37; 95% CI: 1.36–1.39). CONCLUSIONS: Racial/ethnic disparities in hospitalization rate persist among MBs with COVID-19, and associations of COVID-19 hospitalization with chronic conditions differ among racial/ethnic groups in the USA. These findings indicate the need for interventions in racial/ethnic populations at the highest risk of being hospitalized with COVID-19. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1007/s40615-020-00960-y.

The number of cases of SARS CoV-2 infection, the virus that causes COVID-19 illness, has rapidly increased in the USA. Since the first case of patient with COVID-19 was reported in January 2020 [1] , approximately 9.3 million cases and > 232,000 confirmed COVID-19-associated deaths have been reported in the USA as of November 3, 2020 [2] . Studies have reported that COVID-19-associated hospitalizations and mortality in the USA were higher among older adults, males, and Black persons [3, 4] . Risk factors associated with hospital admission include comorbidity, public insurance, residence in a low-income area, and obesity [4] . Persons in any age group with certain underlying medical conditions have been reported to be at increased risk for severe illness from COVID-19; these medical conditions include cancer, chronic kidney disease, chronic obstructive pulmonary disease (COPD), immunocompromised state from solid organ transplant, obesity, serious heart conditions, sickle cell disease, and type 2 diabetes mellitus [5] [6] [7] . Approximately 90% of persons hospitalized for severe illness from COVID-19 have been observed to have one or more of these underlying medical conditions [3] .

Black persons and other racial and ethnic minority groups with COVID-19 also are more likely to have one or more medical conditions than White patients with COVID-19 [8, 9] . Racial and ethnic disparities in COVID-19 hospitalization and mortality have been observed, with higher rates among Black persons, Hispanic persons, and American Indian/ Alaska Native (AIs/ANs) persons, compared with White persons [8, 10] . Although associations of hospitalization with chronic conditions have been reported and concerns regarding racial and ethnic disparities with COVID-19 burden on disproportionately affected populations have been discussed, the association between hospitalized versus nonhospitalized COVID-19 patients with or without chronic conditions has not been well evaluated. Prior studies of COVID-19 hospitalizations did not collect data regarding nonhospitalized patients and did not compare the prevalence of underlying chronic conditions between hospitalized and nonhospitalized patients with COVID-19 [3] . Additionally, assessments have been limited to small samples in certain geographic areas [11] . Therefore, the difference in characteristics between hospitalized and nonhospitalized patients with COVID-19 and the impact of associated chronic conditions has not been assessed in large-scale samples. Limited information is available regarding risk factors, such as social determinants of health, urban versus rural residence, and chronic conditions associated with hospitalization among patients with COVID-19, especially among older adults in different racial and ethnic groups who are likely to have a higher prevalence of certain chronic conditions. Older persons typically have a greater disease and economic burden of disease than younger persons [12] . Medicare claims data for older persons and those with a disability or with end-stage renal disease offer an opportunity for examining the association of racial and ethnic disparities among hospitalized COVID-19 patients with chronic conditions. To fill the knowledge gap, we assessed if Medicare beneficiaries (MBs) with fee-for-service (FFS) claims for COVID-19, who also had a chronic health condition, are more likely to receive hospital inpatient services for COVID-19associated outcomes than COVID-19 MBs without chronic conditions. We also assessed racial and ethnic disparities in hospitalizations among MBs with COVID-19 and chronic conditions.

We used data from the 2020 preliminary monthly Medicare with FFS claims for January 1-September 30, 2020, available through the Centers for Medicare and Medicaid Services (CMS, https://www.cms.gov/) Virtual Research Data Center to calculate hospitalization rates for COVID-19-associated outcomes among MBs. A Medicare COVID-19 confirmed case was defined as an MB with a primary or secondary International Classification of Diseases Clinical Modification (10 rev.) (ICD-10-CM) [13] diagnosis code of B97.29 (other coronavirus as the cause of diseases classified elsewhere) on an FFS claim for January 1-March 31, 2020, or diagnosis code U07.1 (2019 novel coronavirus, COVID-19, confirmed by laboratory testing) from April 1, 2020, onward for any health care setting. A Medicare COVID-19-hospitalized beneficiary was defined as an MB with a diagnosis of COVID-19 on an FFS claim for an inpatient hospital setting.

We classified each COVID case as yes/no for beneficiaries with or without a claim for hospitalization. We estimated frequencies of COVID patients with or without hospitalization by sex (male or female), race/ethnicity (non-Hispanic White [ [14] to the following three urban-rural groups: urban (large central metropolitan and large fringe metropolitan), sub-urban (medium metropolitan and small metropolitan), and rural (micropolitan and noncore). We included 27 chronic health, mental health, and substance abuse conditions as follows: acute myocardial infarction; atrial fibrillation; heart failure; hypertension; ischemic heart disease; stroke; peripheral vascular disease; chronic kidney disease; diabetes; hyperlipidemia; anemia; diabetes; Alzheimer disease; depression; obesity; COPD; schizophrenia; lung cancer; liver disease; viral hepatitis; HIV/AIDS; asthma; osteoporosis; tobacco use; opioid use disorder; spina bifida and other congenital anomalies of the nervous system; and sickle cell disease. We also classified cardiovascular disease, including acute myocardial infarction, atrial fibrillation, heart failure, hypertension, ischemic heart disease, stroke, and peripheral vascular disease. These conditions were defined by using the 2018 (the most recently/updated available data) Chronic Conditions Data Warehouse predefined algorithms [15] . These conditions were selected because of (a) high prevalence among MBs or (b) COVID-19 patients with these conditions likely being at higher risk for hospitalization than those without these conditions. We summarized the number of comorbidities for each beneficiary with or without hospitalization as four categories: 0, 1-3, 4-5, or ≥ 6.

First, we calculated the following two measures: Second, we analyzed the characteristics of FFS MBs with confirmed COVID-19 and with or without COVID-19-related hospitalizations including sex, race and ethnicity, US Census region, urban-rural residence, Medicare entitlement, Medicare eligibility status, and comorbidity. We also examined the differences in characteristics between hospitalized and nonhospitalized COVID-19 MBs in a 2-tailed chi-square test at a significance level of P < .05. Third, we estimated the prevalence of chronic conditions among MBs with a confirmed COVID-19 illness and who had versus who did not have a hospitalization claim. Lastly, we used multivariable logistic regression to examine the association between hospitalization and the selected chronic conditions among patients with COVID-19 by race/ethnicity. We estimated adjusted odds ratios (aORs) with 95% confidence intervals (95% CIs) for COVID-19 hospitalization among those COVID-19 MBs with each chronic condition versus those COVID-19 MBs without that chronic condition, adjusting for age, sex, and urban-rural residence. We used a 2-tailed Wald chi-square test for statistical significance at P < .05 for differences in the rate of hospitalized beneficiaries with COVID-19 and with a chronic condition versus those COVID-19 MBs without that chronic condition. All analyses were performed by using SAS® 9.4 and SAS Enterprise Guide® 7.1 (SAS Institute, Inc., Cary, NC) in the secured environment of the CMS Virtual Research Data Center through the Chronic Conditions Data Warehouse [16] . Federal officials not involved in the study approved the ethical conduct of the analysis and reporting of deidentified Medicare data that did not involve human research.

Of 39.3 million MBs with FFS claims during January 1-September 30, 2020, the total number of MBs with COVID-19 was 710,980; among those, 207,600 COVID-19-related hospitalizations occurred as of September 30, 2020. The hospitalization rate for COVID-19 was 527.9/100,000 MBs ( Table 1 Table 2 ). Supplemental Table 1 presents the prevalence of chronic conditions among MBs with a confirmed COVID-19 illness with hospitalization versus without hospitalization. The 10 chronic conditions ranked for highest MBs with COVID-19 infection who were hospitalized versus nonhospitalized were observed for chronic kidney disease (11.98%), diabetes (9.68%), and heart failure (7.35%). 

Our findings indicate that racial and ethnic disparities exist in hospitalization rates among enrolled MBs and in MBs with COVID-19. Racial and ethnic minorities (e.g., NHBs and AI/ ANs) were more likely to be hospitalized than NHWs. Especially, AI/AN MBs with COVID-19 were more likely to require inpatient medical services than patients in any other racial and ethnic groups after becoming infected. Our findings also indicate that substantial variation exists between hospitalized and nonhospitalized MBs with COVID-19 by demographic attributes, including age, sex, race/ethnicity, geographic region, urbanization, Medicare entitlement, Medicare, and Medicaid dual-eligibility status, and comorbidity. Hospitalized MBs with COVID-19 were more likely to be aged ≥ 75 years, to be male, to be Black, to reside in the northeast US region, to reside in urban counties, to have end-stage renal disease, to be enrolled in both Medicare and Medicaid programs, and to have multiple chronic conditions. Social determinants of health are known to play a principal role in disease outcomes and have a substantial effect on life expectancy [17] . They also seem to have a considerable effect on COVID-19 morbidity and mortality and hospitalizations for COVID-19-associated outcomes [18, 19] . Inequities in the social determinants of health that disproportionately affect racial and ethnic minorities include poverty, lack of health care access to testing centers and medical facilities, being uninsured, using public transportation, employees in jobs that require physical contact, or living in crowded spaces and housing environments [20, 21] . A prior study reported that the COVID-19 illness rate is three times higher among predominantly Black counties than among predominantly White counties in the USA [22] . Similarly, our results also found that the COVID-19-associated hospitalizations for NHB MBs were 3 times or higher than NHW MBs. Further studies regarding social inequalities in health are needed to quantify the effect on COVID-19 patients among all disproportionately affected populations in different age and racial/ethnic groups.

Our results support prior analyses [3] [4] [5] [6] [7] in that they demonstrate a statistically significant association (P < .0001) between hospitalizations attributable to COVID-19 illness and selected chronic conditions, including end-stage renal disease, chronic kidney disease, cardiovascular disease, COPD, diabetes, obesity, hypertension, and lung cancer. Medicare MBs infected with COVID-19, who have one or more of the selected chronic conditions, were at increased risk for being admitted to the hospital for illness from COVID-19 compared with those without these chronic conditions. We also observed that the statistical significance of associations of hospitalization and these chronic conditions varied slightly among the different racial and ethnic groups. For example, end-stage renal disease, heart-related conditions, COPD, lung cancer, opioid use disorder, stroke, and spina bifida and other congenital anomalies of the nervous system were significantly associated with hospitalizations among NHW MBs with COVID-19; sickle cell disease among NHB MBs; hepatitis among Hispanic MBs; and obesity among A/PI MBs. Although the magnitude of the risk varied by chronic condition and by race and ethnicity, COVID-19 MBs with one of these chronic conditions were at increased risk to be hospitalized for severe outcomes from COVID-19 than those without. The existing and persistent racial and ethnic disparities among US communities have posed considerable challenges for public health interventions during the US COVID-19 pandemic in 2020.

Our results indicate that the prevalence of such conditions was higher among hospitalized MBs compared with nonhospitalized patients (e.g., chronic kidney disease, diabetes, heart failure, ischemic heart disease, and hypertension). These selected chronic conditions were more prevalent (e.g., a difference of 11.98% for chronic kidney disease) among hospitalized versus nonhospitalized MBs with COVID-19. Our analysis represents estimates for the Medicare population with a large proportion of older adults; thus, prevalence of chronic conditions in our study might be higher than it is among the general population.

For example, a comprehensive systematic review using meta-analyses reported approximately 16%, 12%, 8%, and 8% prevalence rates among hospitalized patients for hypertension, cardiovascular disease, smoking history, and diabetes, respectively [23] . A recent CDC report from COVID-NET data stated that approximately 90% of persons hospitalized have one or more underlying medical conditions [3] .

Our findings are subject to several limitations. First, these results were based on MBs who had claims for COVID-19 illness and claims for hospitalization from COVID-19 illness. We were unable to obtain characteristics of MBs who were asymptomatic, were false positive, and might not have a claim for COVID-19 in the Medicare data system. Second, our data are claims-based, and the population is restricted to MBs with FFS claims covered by Medicare insurance; claims were submitted by physicians' offices, inpatient hospitals, or laboratories. Our estimates cannot account for claims not submitted from health care settings. Additionally, demographic distributions and chronic disease prevalence among the Medicare population are different from the US general population. Therefore, our estimates might not be directly comparable with those in prior studies that used data collected from surveillance systems. Third, this analysis includes preliminary data from the monthly Medicare FFS claims only but did not include Medicare Advantage encounter data because those 2020 data are not yet available. Moreover, these results are based on the preliminary monthly inpatient and outpatient data, which might change as additional data are available. Historically, only approximately 40% of FFS claims are received by CMS during the first month after the date of service, and 90% are received within 3 months [24] . Our estimates are likely to represent an underestimate because of the delay in data collection and data reporting. Fourth, chronic condition prevalence in this study was calculated by using conditionspecific variables from the most recent 2018 Chronic Conditions Data Warehouse. Our estimates in chronic condition prevalence might be changed if more current chronic condition data become available. Fifth, we estimated the association of COVID-19 hospitalization with each chronic condition separately but did not adjust for the combined effect of multiple chronic conditions. Sixth, our analysis used ICD-10-CM codes defined by CMS [24] . Therefore, our results might not be directly comparable with estimates for cases defined by other sources [25] . Lastly, we assessed hospitalizations among MBs with COVID-19 with chronic conditions by using limited data and information (e.g., basic demographic characteristics) available in Medicare claims data. Other factors contributing to health care usage, racial and ethnic disparities, or possible associated conditions for illness were not considered in the analyses (e.g., physician availability, geographical access to services, education, or local policies in preventive measures, including social distancing or wearing face coverings), or other conditions (e.g., pregnancy).

Our findings confirm the association between hospitalization for COVID-19 illness and chronic conditions among MBs with COVID-19 infection. Racial/ethnic disparities in hospitalization rate persist among MBs with COVID-19, and associations of COVID-19-related hospitalization with chronic conditions differ among racial/ethnic groups in the USA. These findings indicate the need for interventions in racial/ ethnic populations at the highest risk of being hospitalized with COVID-19. The analysis and findings of the study illustrate how Medicare claims data can be used within limitations to supplement data available from notifiable disease surveillance systems and hospital utilization surveys.

Supplementary Information The online version contains supplementary material available at https://doi.org/10.1007/s40615-020-00960-y.

et al; China Medical Treatment Expert Group for Covid-19. Clinical characteristics of coronavirus disease 2019 in China

Hospitalization rates and characteristics of patients hospitalized with laboratory-confirmed coronavirus disease 2019-COVID-NET, 14 states

Hospitalization and mortality among black patients and white patients with COVID-19

Coronavirus disease 2019 (COVID-19)-people with certain medical conditions

Comorbidity and its impact on 1590 patients with COVID-19 in China: a nationwide analysis

Does comorbidity increase the risk of patients with COVID-19: evidence from meta-analysis

Race/ethnicity, underlying medical conditions, homelessness, and hospitalization status of adult patients with COVID-19 at an urban safety-net medical center

COVID-19 and racial/ ethnic disparities

Characteristics of persons who died with COVID-19-United States

Characteristics associated with hospitalization among patients with COVID-19

International Classification of Diseases, Tenth Revision, Clinical Modification (ICD-10-CM)

NCHS urban-rural classification scheme for counties. National Center for Health Statistics

Chronic Conditions Data Warehouse. Condition categories

Chronic Conditions Data Warehouse. Data dictionaries

Social determinants of health in the United States: addressing major health inequality trends for the nation

COVID-19 and the impact of social determinants of health

The detrimental impact of the COVID-19 crisis on health equity and social determinants of health

COVID-19: a potential public health problem for homeless populations

COVID-19 and health equity-a new kind of "herd immunity

COVID-19 and African Americans

Prevalence of underlying diseases in hospitalized patients with COVID-19: a systematic review and meta-analysis

Preliminary Medicare COVID-19 data snapshot

Centers for Disease Control and Prevention. ICD-10-CM official coding and reporting guidelines

Publisher's Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations

Acknowledgments The authors thank the CMS staff for their data support and disclosure view, staff in the Chronic Conditions Warehouse Help Desk for software and technical support, and staff in the Centers for Disease Control and Prevention's Center for Surveillance, Epidemiology, and Laboratory Services for their administrative support. We extend special thanks to C. Kay Smith, MEd, CDC, for editorial support.Author Contributions Chang had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.Study 

Conflict of Interest The authors declare that they have no conflict of interest.Ethics Approval Federal officials not involved in the study approved the ethical conduct of the analysis and reporting of deidentified Medicare data that did not involve human research.Disclaimer The findings and conclusions in this report are those of the authors and do not necessarily represent the official position of the Centers for Disease Control and Prevention.