Characteristics and Factors Associated With Coronavirus Disease 2019 Infection, Hospitalization, and Mortality Across Race and Ethnicity

Abstract

Background

Data on the characteristics of coronavirus disease 2019 (COVID-19) patients disaggregated by race/ethnicity remains limited. We evaluated the sociodemographic and clinical characteristics of patients across racial/ethnic groups and assessed their associations with COVID-19 outcomes.

Methods

This retrospective cohort study examined 629 953 patients tested for severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) in a large health system spanning California, Oregon, and Washington between March 1 and December 31, 2020. Sociodemographic and clinical characteristics were obtained from electronic health records. Odds of SARS-CoV-2 infection, COVID-19 hospitalization, and in-hospital death were assessed with multivariate logistic regression.

Results

A total of 570 298 patients with known race/ethnicity were tested for SARS-CoV-2, of whom 27.8% were non-White minorities: 54 645 individuals tested positive, with minorities representing 50.1%. Hispanics represented 34.3% of infections but only 13.4% of tests. Although generally younger than White patients, Hispanics had higher rates of diabetes but fewer other comorbidities. A total of 8536 patients were hospitalized and 1246 died, of whom 56.1% and 54.4% were non-White, respectively. Racial/ethnic distributions of outcomes across the health system tracked with state-level statistics. Increased odds of testing positive and hospitalization were associated with all minority races/ethnicities. Hispanic patients also exhibited increased morbidity, and Hispanic race/ethnicity was associated with in-hospital mortality (odds ratio [OR], 1.39; 95% confidence interval [CI], 1.14–1.70).

Conclusion

Major healthcare disparities were evident, especially among Hispanics who tested positive at a higher rate, required excess hospitalization and mechanical ventilation, and had higher odds of in-hospital mortality despite younger age. Targeted, culturally responsive interventions and equitable vaccine development and distribution are needed to address the increased risk of poorer COVID-19 outcomes among minority populations.

Since the coronavirus disease 2019 (COVID-19) was first reported in Washington state, the United States has documented the highest number of confirmed cases and deaths in the world. Increasing evidence has indicated that COVID-19 disproportionately affects patients of minority race and ethnicity [1–3]. Although reports have identified different rates of infection, hospitalization, and mortality among minority populations, there is limited information on the characteristics of COVID-19 patients disaggregated by race/ethnicity [4, 5].

The prevalence of comorbidities and social environments vary between racial/ethnic groups [6, 7]. Some of these characteristics, including obesity and crowded housing, are potential risk factors for COVID-19 and disease severity [8, 9]. Therefore, understanding how the characteristics of patients differ between races/ethnicities and which factors are associated with disease outcomes are critical for public health and designing community-based interventions. Unfortunately, such detailed characteristics remain sparse, and certain racial/ethnic groups, specifically Asian Americans, Native Hawaiians/Pacific Islanders (NH/PI), and American Indians/Alaska Natives (AI/AN) remain yet to be characterized in detail [10, 11]. Furthermore, although sociodemographic and health characteristics vary across geography, multistate comparisons are limited. The objective of this study therefore is to examine the characteristics of and factors associated with severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection, hospitalization with COVID-19, and in-hospital mortality in a large diverse population of patients in a large health system operating in California, Oregon, and Washington.

METHOD

Study Design, Setting, and Population

This retrospective cohort study included patients from California, Oregon, and Washington who were tested for SARS-CoV-2 with a polymerase chain reaction assay of a nasopharyngeal sample and were seen at a Providence St. Joseph Health (PSJH) facility between March 1, 2020, and December 31, 2020. In-hospital outcomes were monitored through January 31, 2020. PSJH is one of the largest health systems in the United States. In 2019, approximately 3.5 million patients in California, Oregon, and Washington received care at the facilities included in this study, of whom 62.5% identified as non-Hispanic White, 11.5% as Hispanic, 6.8% as non-Hispanic Asian American, 3.6 % as non-Hispanic Black, 0.6% non-Hispanic AI/AN, 0.5% non-Hispanic NH/PI, 5.7% non-Hispanic other, and 8.8% unknown. The protocol for this study was approved by the PSJH institutional review board (#STUDY2020000203).

Data Collection

Patient demographic and clinical data were extracted from PSJH’s Epic electronic health record system. Patients with a positive polymerase chain reaction test for SARS-CoV-2 were considered to have a confirmed SARS-CoV-2 infection. For patients who had multiple tests, only the initial positive test result was considered. Extracted demographic data included age, sex, race, ethnicity, and insurance plan. One hundred thirty eight patients with missing sex were excluded from the study. Missing race or ethnicity was grouped as unknown. ZIP codes were used to identify the neighborhood-level median income, crowded housing (> 1 person per room), minority population (race/ethnicity except non-Hispanic white), and limited English-proficient speakers from the US Census Bureau’s American Community Survey. Clinical data include underlying medical conditions identified using ICD-10-CM codes or direct clinical measurements linked to past encounters between January 1, 2019, and the date of SARS-CoV-2 testing. We included underlying medical conditions that have previously been associated with COVID-19 [12, 13]. Charlson Comorbidity Index (CCI) was used to capture the risk from multiple comorbidities. Obesity and hypertension, which are not part of CCI, were also included. We used previously defined diagnosis codes for components of CCI [14]; code I10 was used for hypertension and body mass index classified based on Centers for Disease Control and Prevention definitions for obesity. Inpatient encounter data included presenting vital signs, baseline laboratories, supplemental oxygen use, length of stay, transfer to intensive care unit, and discharge disposition.

For comparison with state-level data, COVID-19 cases and deaths for California, Oregon, and Washington was obtained from the COVID Racial Data Tracker, which aggregates historical data from state agencies [15]. Data on COVID-19 hospitalization for California and Oregon were obtained from the Centers for Disease Control and Prevention’s COVID-NET, whereas Washington data were obtained from the COVID Racial Data Tracker [15, 16].

Statistical Analysis

We compared the sociodemographic and clinical characteristics of patients across COVID-19 outcomes and race/ethnicity categories, defined as Hispanic, non-Hispanic Black (Black), non-Hispanic Asian American (Asian), non-Hispanic NH/PI; non-Hispanic AI/AN, non-Hispanic white (White), and non-Hispanic other (other), which includes multirace/ethnicity. For COVID-19 hospitalized patients, presenting clinical characteristics were available and included clinical status on the World Health Organization (WHO) 9-point Clinical Progression Scale; [17] presenting vitals within the first 6 hours of admissions; and baseline laboratory test results within 24 hours of admissions. The WHO Clinical Progression Scale was developed to measure clinical illness of an COVID-19 infection and consists of the following categories: 0, uninfected; 1, ambulatory, no limitation of activity; 2, ambulatory, limitation of activity; 3, hospitalized, no oxygen therapy; 4, hospitalized, oxygen by mask or nasal prongs; 5, hospitalized, noninvasive ventilation or high-flow oxygen; 6, hospitalized, intubation and mechanical ventilation; 7, hospitalized, ventilation + additional organ support; and 8, death.

Associations with SARS-CoV-2 infection, COVID-19 hospitalization, and in-hospital mortality was assessed using mixed-effect logistic regression models with state and month of diagnosis as random effect variables to account for geographic and temporal variations. For each outcome, we fitted both unadjusted univariate models and adjusted multivariate models. All multivariate models included race/ethnicity as an independent variable, with demographic factors (age, age squared, sex), socioeconomic factors (insurance, neighborhood median income, crowded housing, limited English proficiency, and minority), and comorbidities (CCI, hypertension, obesity) as covariates. An age-squared term was included in addition to age to capture the nonlinear relationship between COVID-19 outcomes and age [18]. For analyses of hospital mortality, additional covariates included presenting WHO Clinical Progression Scale score and baseline laboratory results. Covariates were selected based on previously identified risk factors and patterns of missingness and collinearity. Certain characteristics—body mass index, insurance coverage, and baseline laboratory results—were not available for all patients. Alanine transaminase was excluded because of high correlation with aspartate transaminase and had higher missingness. For variables with less than 20% missingness, missing values were imputed with multiple imputation by fully conditional specification (15 imputations). Statistical analyses were performed with R v3.6.2.

RESULTS

Characteristics of Patients Tested for SARS-CoV-2

A total of 629 953 patients tested for SARS-CoV-2 were included. 570 298 patients (90.5%) reported race/ethnicity, of which 72.2% were White, 13.4% were Hispanic, 5.4% were Asian, 3.8% Black, 0.9% were AI/AN, 0.6% were NH/PI, and 3.7% were other (Table 1). The mean age among all tested patients was 51.5 ± 19.4 years and 57.0% were female. The majority of patients (53.7%) had public insurance (Medicaid, 30.0%; Medicare, 23.7%), with higher percentages of Hispanic, Black, and AI/AN patients on Medicaid than White, Asian, and NH/PI patients (Table S1). The most common comorbidities were obesity (37.1%), hypertension (23.3%), diabetes (9.4%), and asthma (6.5%). The median score on the CCI was 1.0 (95% CI, 0.0–3.0).

Characteristics of Patients Positive for SARS-CoV-2 Infection

Of the 629 953 patients tested for SARS-CoV-2, 54 645 (8.7%) were positive. Among the 49 081 patients with known race/ethnicity, the rate of positive test results was higher for minority patients than White patients (5.9%; Table S1). Hispanic and NH/PI patients had the highest rates (22.1% and 13.5%, respectively). Consequently, the racial/ethnic composition of SARS-CoV-2–infected patients were 49.9% White, 34.3% Hispanic, 5.0% Asian, 4.8% Other, 4.2% Black, 1.0% NH/PI (1.4%), and 0.9% AI/AN patients (Table 1).

Among all SARS-CoV-2–infected patients, the mean age was 47.8 ± 19.2 years and 52.6% were female (Table 1). Compared with White patients, mean ages were lower among patients of minority race/ethnicity, except for Asians, which had a similar mean age (Table 2). CCI scores were also lower among minority patients. The prevalence of diabetes, however, was higher among minority patients. Additionally, relative to White patients, Hispanic, Black, NH/PI, and AI/AN patients had higher prevalence of obesity; Asian, Black, NH/PI, and AI/AN patients had higher prevalence of hypertension; and Black, NH/PI, and AI/AN patients had higher prevalence of both asthma and kidney disease.

Hispanic, Black, NH/PI, and AI/NH patients were more likely to have Medicaid insurance than were White and Asian patients (Table 2). All minority patients were more likely than White patients to reside in neighborhoods with higher percentages of crowded housing and minorities. Hispanic, Asian, and Black patients in particular were also more likely to live in neighborhoods with a higher percentage of limited English-proficient population. All minority patients were more likely to have received their tests in the emergency department than White patients.

Characteristics of Patients Hospitalized With COVID-19

A total of 15.6% (n = 8536) of the patients who tested positive for COVID-19 were hospitalized. 8210 (96.2%) had known race/ethnicity, of which 45.7% were White, 35.3% were Hispanic, 6.5% were Asian, 4.5% were Black, 1.2% were NH/PI, 0.9% were AI/AN, and 5.9% were other (Table 1). The mean age of all patients was 64.7 ± 17.6 years and 54.5% were male. The median CCI score was 3.0 (2.0–6.0) and the most common comorbidities were obesity (42.4%), hypertension (40.2%), and diabetes (28.3%).

White patients had the highest mean age (70.5 ± 16.2 years; Table 3) and median CCI score (4.0 [3.0–7.0]). NI/PI patients had the lowest mean age (55.9 ± 17.3 years), and lowest median CCI score (2.0 [1.0–5.0]). The prevalence of obesity was highest among AI/AN, NH/PI, and Hispanic patients and lowest among Asian patients. However, Hispanics had the lowest prevalence of hypertension, whereas Asian and Black patients had the highest. The majority of Hispanic, Black, and AI/AN patients were on Medicaid; and Hispanic, Black, and Asian patients resided in neighborhoods with higher percentages of crowded housing, minorities, and limited English speakers than other racial/ethnic subgroups.

On admission, a higher percentage of Hispanic patients (11.1%) than White patients (6.7%) had a score of 5 or above on the WHO Clinical Progression Scale, and a higher percentage of Hispanic patients than White patients were febrile, had low oxygen saturation, and had high respiration rates. Over the course of hospitalization, a higher percentage of Hispanic patients (18.5%) than White patients (11.1%) needed mechanical ventilation.

Characteristics of SARS-CoV-2–infected Patients Over Time

Throughout the pandemic, rates of positive test, hospitalization, and in-hospital mortality per 100 000 patients were on average higher for Black, Hispanic, NH/PI, and AI/AN patients than Asian and White patients (Figure 1). Hispanic patients, in particular, had the highest rates, especially during the June–July and November–December resurgence of COVID-19. The mean age of SARS-CoV-2–infected patients generally decreased over time, whereas the mean age of COVID-19 hospitalized patients and patients who experienced in-hospital mortality remained relatively consistent.

Comparison With State-level SARS-CoV-2 Cases, COVID-19 Hospitalization, and Death

The distribution of race/ethnicity among patients in this study generally reflected state-level distributions for COVID-19 outcomes (Figures S1–3), including for mortality in which state-level data captured both in-hospital and out-of-hospital deaths. High proportions of Hispanics were consistently observed in both the statewide data and the health system. The differences in racial/ethnic distributions between the health system population and the catchment population were consistent or smaller than the differences observed at the state level (Figures S4–5).

Factors Associated With SARS-CoV-2 Infection

Minority populations including Hispanic, Black, Asian, NH/PI, and AI/AN had increased odds of SARS-CoV-2 infection compared with Whites in unadjusted and adjusted analysis (Table 4). In adjusted multivariate analysis, increased odds of positive SARS-CoV-2 results were independently associated with Hispanic (OR, 3.09; 95% CI, 2.99–3.18), NH/PI (OR, 2.23; 95% CI, 2.01–2.48), Black (OR, 1.35; 95% CI, 1.28–1.43), Asian (OR, 1.31; 95% CI, 1.25–1.38), and other (OR, 1.69; 95% CI, 1.6–1.78) race/ethnicity, with White patients as the reference category. Increasing age and age squared, male sex, overweight, obesity (all categories), Medicaid insurance, lack of insurance, and residence in a neighborhood with higher percentage limited English-proficient individuals were also independently associated with increased odds of infection. Higher CCI score, Medicare insurance, and higher median income, however, were associated with lower odds of positive SARS-CoV-2 infection.

Factors Associated With Hospitalization for COVID-19

Minority races/ethnicities were also associated with increased odds for COVID-19 hospitalization (Table 4). In multivariate analysis, the OR for COVID-19 hospitalization was highest among NH/PI patients (OR, 2.01; 95% CI, 1.55–2.61]), followed by Asian (OR, 1.62; 95% CI, 1.43–1.84), AI/AN (OR, 1.56; 95% CI, 1.17–2.06), other (OR, 1.32; 95% CI, 1.16–1.5), Hispanic (OR, 1.31; 95% CI, 1.22–1.42), and Black (OR, 1.18; 95% CI, 1.02–1.36) patients. Increasing age, male sex, public insurance, no insurance, higher CCI score, being underweight, having class 2 or 3 obesity, having hypertension, and residence in a neighborhood with higher rates of limited English proficiency were also independently associated with increased odds of hospitalization.

Factors Associated With Hospital Mortality in Admissions for COVID-19

In adjusted multivariate analysis, Hispanic race/ethnicity was significantly associated with increased odds of in-hospital mortality (OR, 1.41; 95% CI, 1.15–1.71). Other minority races/ethnicities, however, were not significantly associated (Table 4). Hospital mortality was also independently associated with age; class 3 obesity; public insurance; higher score on the WHO Clinical Progression Scale and CCI; and high white blood cell count, low lymphocyte, low platelet count, high aspartate transaminase, high blood urea nitrogen, high creatinine, and high bilirubin. Interaction analysis of race/ethnicity and age further identified disproportionately increased odds of hospital mortality among Hispanic patients as age increased (OR, 1.30; 95% CI, 1.04–1.62]; P_interaction = 0.021; Figure S6).

DISCUSSION

This study examined the characteristics and clinical outcomes of 629 953 patients tested for SARS-CoV-2 across California, Oregon, and Washington. Overall, we highlight how characteristics of SARS-CoV-2–infected patients vary by race/ethnicity and show differential associations of COVID-19 hospitalizations, morbidity, and mortality across race/ethnic subpopulations. Although patients of minority race/ethnicity represented 27.8% of patients tested for SARS-CoV-2, they constituted 50.1% of SARS-CoV-2–infected patients and 54.3% of COVID-19 hospitalized patients. Hispanic patients in particular represented 34.3% of SARS-CoV-2 infected and 35.3% of hospitalized patients despite making up only 13.4% of tested patients, a pattern consistent with state-level statistics. Race/ethnicity was associated with SARS-CoV-2 infection and COVID-19 hospitalization, with increased odds of both outcomes highest among Hispanics and NH/PI patients. Hispanic race/ethnicity was also significantly associated with increased odds of hospital mortality. These findings showcase the disproportionate burden of COVID-19 born by Hispanics in these western US states, despite being of younger age and having generally fewer comorbidities than White patients.

Understanding racial and ethnic disparities in COVID-19 cases and outcomes is important for understanding the nature of the disease and guiding public health prevention efforts and medical interventions. This study comprehensively detailed and compared the clinical and epidemiological characteristics of COVID-19 across all major races/ethnicities, highlighting in particular Asian, NH/PI, and AI/AN patients who have been underrepresented in COVID-19 literature to date. In doing so, we show that these populations, particularly NH/PI patients, have similar or higher odds of SARS-CoV-2 infection and hospitalization as Black and Hispanic patients. Concomitantly, we show that factors associated with COVID-19 severity vary across race/ethnicity. These findings demonstrate the need to further focus resources on addressing COVID-19 in all minority communities through efforts such as culturally appropriate public health messaging, data collection, improving access to testing, and possibly actively intervene earlier in the disease course [19].

The significant associations of minority races/ethnicities with SARS-CoV-2 infection and COVID-19 hospitalization builds on previous analyses of Black and Hispanic patients [20–23]. However, unlike previous studies, we also found a significant association between Hispanic race/ethnicity and hospital mortality and a significant interaction between Hispanic race/ethnicity and age that signifies the relationship between Hispanic race/ethnicity and hospital mortality is moderated by age. The clinical characteristics of Hispanic patients at hospital admission suggest that they are presenting with more severe illness than White patients because Hispanic patients are more likely to be febrile, have low oxygen saturation, and high respiration rates. The higher percentage of Hispanic patients presented with WHO Clinical Progression Scale scores of 5 or higher reflects a need for high-flow supplemental oxygen or mechanical ventilation. These clinical findings indicate a delay in seeking care among Hispanic patients. Although this study cannot identify the causes behind the observed associations, certain social, structural, or biologic determinants of health have been suggested [3]. Social and structural determinants could include occupation risk and limited access to healthcare and testing [3]. Additional studies are needed to identify the causal factors driving disparities in COVID-19.

Racial/ethnic disparities continue to persist as the pandemic endures. Our temporal analysis revealed that resurgences of COVID-19 particularly burdened minority patients. As vaccines and treatments against COVID-19 are studied, meaningful representation of diverse populations in clinical trials, particularly of those disproportionately affected by COVID-19, is necessary to ensure their safety and efficacy [24]. Such representation can also help build public trust in the studied vaccine or treatment. Distribution of vaccines should also take into consideration racial equity because it can help mitigate the disproportionate impact of the virus and prevent widening health disparities [25]. Unfortunately, minority populations face various barriers to vaccination, including lack of access to healthcare, distrust of the healthcare system, communication barriers, and misinformation. Thus, additional resources should be placed on reaching high-risk populations, building trust, and reducing barriers to vaccination among minorities.

Although the large size of this study’s diverse cohort and its multistate distribution are strengths of this study, there are limitations. This study was limited to a single health system and certain catchment areas within California, Oregon, and Washington. Thus, the results may be less generalizable to other regions. In particular, the racial/ethnicity composition of COVID-19 patients in our study may not reflect those in other states and at the national level. For example, Black patients are underrepresented in our study when compared with national-level statistics. At the same time, electronic health records are subject to the quality, consistency, and completeness of entry by providers. Some patients may have also received care at other institutions and therefore certain outcomes or characteristics may be underreported. Our use of International Classification of Diseases, 10th Revision, Clinical Modification code may not capture diagnoses that are not billable or represented by a given code. Despite these limitations, however, our results highlight how the impact of COVID-19 vary across race/ethnicity in a large geographical area.

Notes

Author contributions. Concept and design: C. L. D., S. A. K., R. T. R., G. D., J. D. G., and J. J. H. Acquisition, analysis, or interpretation of data: C. L. D., S. A. K., R. T. R., G. D., J. D. G., J. J. H., and A. T. M. Drafting of the manuscript: C. L. D. Critical revision of the manuscript for important intellectual content: all authors. Statistical analysis: C. L. D. and S. A. K. Obtained funding: J. R. H., J. D. G., J. J. H., and A. T. M. Supervision: J. J. H. and A. T. M. C. L. D., S. A. K., and R. T. R. had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Acknowledgments. The authors thank the caregivers in the Providence St. Joseph Health (PSJH) system for their tireless efforts in caring for patients with coronavirus disease 2019. They also thank Mark Premo, Lindsay Mico, Jennifer Jones, Andrey Dubovoy, Vivek Tomer, and the PSJH Health Intelligence team for their support and management of the electronic health record data. Data used in this study is archived within PSJH in a Health Insurance Portability and Accountability Act–secure location to facilitate verification of study conclusions.

Disclaimer. No funders had a role in designing the study; collecting, analyzing, or interpreting the data; or preparing, reviewing, or approving the manuscript for submission/publication.

Financial support. This work was supported by Biomedical Advanced Research and Development Authority under contract HHSO10201600031C administered by Merck and Co. J. J. H. and R. T. R. were also funded by National Center for Advancing Translational Sciences of the National Institutes of Health under award number OT2 TR003443.

Potential conflicts of interest. J. R. H. reports fees and support from PACT Pharma, Isoplexis, Indi Molecular; advisory board member fees from Nanostring; consultant fees from Merck; and SAB member fees and equity from Atlasxomics outside the submitted work. J. D. G. performed contracted research with Gilead Sciences, Regeneron Pharmaceuticals and Eli Lilly and has received grants from Monogram Biosciences and Viracor outside of the submitted work. G. D. reports personal fees from Gilead Sciences, consulting fees from Lexicon Pharmaceuticals, scientific advisory board fees from Safeology, Inc, and Clinical trial research support from Gilead Sciences, Regeneron Inc, Roche Inc, Boehringer Ingelheim, and Edesa Biotech, outside the submitted work. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

References

Author notes