key: cord-0857718-7hnh85wy
authors: Mendy, A.; Apewokin, S.; Wells, A. A.; Morrow, A. L.
title: Factors Associated with Hospitalization and Disease Severity in a Racially and Ethnically Diverse Population of COVID-19 Patients
date: 2020-06-26
journal: medRxiv : the preprint server for health sciences
DOI: 10.1101/2020.06.25.20137323
sha: 27caae90d0bfbed511d9c60db202438f2bec7c60
doc_id: 857718
cord_uid: 7hnh85wy

Background: The coronavirus disease (COVID-19) first identified in Wuhan in December 2019 became a pandemic within a few months of its discovery. The impact of COVID-19 is due to both its rapid spread and its severity, but the determinants of severity have not been fully delineated. Objective: Identify factors associated with hospitalization and disease severity in a racially and ethnically diverse cohort of COVID-19 patients. Methods: We analyzed data from COVID-19 patients diagnosed at the University of Cincinnati health system from March 13, 2020 to May 31, 2020. Severe COVID-19 was defined as admission to intensive care unit or death. Logistic regression modeling adjusted for covariates was used to identify the factors associated with hospitalization and severe COVID-19. Results: Among the 689 COVID-19 patients included in our study, 29.2% were non-Hispanic White, 25.5% were non-Hispanic Black, 32.5% were Hispanic, and 12.8% were of other race/ethnicity. About 31.3% of patients were hospitalized and 13.2% had severe disease. In adjusted analyses, the sociodemographic factors associated with hospitalization and/or disease severity included older age, non-Hispanic Black or Hispanic race/ethnicity (compared non-Hispanic White), and smoking. The following comorbidities: diabetes, hypercholesterolemia, asthma, chronic obstructive pulmonary disease (COPD), chronic kidney disease, cardiovascular diseases, osteoarthritis, and vitamin D deficiency, were associated with hospitalization and/or disease severity. Hematological disorders such as anemia, coagulation disorders, and thrombocytopenia were associated with higher odds of both hospitalization and disease severity. Conclusion: This study confirms race and ethnicity as predictors of severe COVID-19 and identifies clinical risk factors not previously reported such a vitamin D deficiency, hypercholesterolemia, osteoarthritis, and anemia.

In December 2019, an outbreak of severe acute respiratory syndrome coronavirus-2 (SARS-CoV-2) pneumonia appeared in Wuhan and rapidly spread throughout the world causing more than 8 million infections and close to 450,000 deaths by June 2020. 1, 2 Similar to previous coronaviruses, the newly identified virus is highly contagious; it is transmitted primarily through droplet spread and causes major outbreaks in the absence of adequate control measures. 3 The manifestations of the coronavirus disease 2019 (COVID-19) caused by SARS-CoV-2 are widely variable. The infection is asymptomatic in some individuals, while in others, it causes symptoms ranging from dry cough and dyspnea to severe pneumonia with respiratory failure requiring admission in intensive care unit (ICU) and leading to death in severe cases. 4, 5 To date, a number of studies have investigated the predictors of disease severity in COVID-19. Nevertheless, the sociodemographic and clinical factors that influence this disease course have not been fully defined. [6] [7] [8] [9] [10] [11] [12] Whether asthma, a common respiratory illness affecting 1 in 12 American adults, is associated with disease severity in COVID-19 is still a matter of debate. [13] [14] [15] [16] [17] Other important clinical factors remain understudied. For example, vitamin D deficiency affects 1 in 4 American adults, has an important immunologic role and has been suggested to increase the risk of SARS-CoV-2 infection. 18 However, vitamin D deficiency has not been analyzed as a risk factor for severe COVID-19 among patients infected with the virus.

Another understudied clinical factor is hypercholesterolemia, which affects 1 in 8 American adults; whether this condition has been reported to be associated with severe COVID-19 has not been well-analyzed in the published literature. 19, 20 Moreover, most of the research on the potential factors for severe COVID-19 have been conducted in China among Chinese patients All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) 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 June 26, 2020. . https://doi.org/10.1101/2020.06. 25.20137323 doi: medRxiv preprint which may limits their generalizability to multi-ethnic populations. [6] [7] [8] 10 and only a few reports have included a racially diverse study sample. 9, 11, 12, 21 To address this gap in the literature, we aimed to determine factors associated with hospitalization and disease severity in a racially and ethnically diverse population of COVID-19 patients and including some comorbidities missing in previous studies.

Data were extracted from the electronic medical record system for all COVID-19 patients diagnosed at the University of Cincinnati health system (UC Health) between March 13, 2020 to May 31, 2020. UC Health consists of 4 hospitals located in the Cincinnati metropolitan area and primary care and specialty clinics located in the states of Ohio, Kentucky, and Indiana. COVID-19 diagnosis was defined as a positive nasopharyngeal reverse transcriptase polymerase chain reaction test for SARS-CoV-2. A total of 691 patients were diagnosed with COVID-19 and after exclusion of 2 patients with no data on sex, 689 patients were included in our study.

Age at the time of COVID-19 diagnosis was calculated using patients date of birth. Sex, race/ethnicity, and smoking were self-reported. Comorbidities, as well as hematological disorders were defined using the 10 th revision of the International Classification of Diseases (ICD10) codes. The following comorbidities were well characterized in our data and were All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) 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 June 26, 2020. . https://doi.org/10.1101/2020.06. 25.20137323 doi: medRxiv preprint included in the study: obesity (E66), diabetes (E10 and E11), pure hypercholesterolemia (E78.0), asthma (J45), chronic obstructive pulmonary disease (COPD) (J44), chronic kidney disease (N18), cardiovascular disease (I00-I99), neoplasm or a history of neoplasm (C00-D49), osteoarthritis (M15-M19), and vitamin D deficiency (E55). The hematological disorders we included were anemia (D50-D53 for nutritional anemia and D55-D59 for hemolytic anemia), coagulation defects, purpura and other hemorrhagic conditions (D65-D69), and thrombocytopenia (D69.6).

Hospitalization was defined as admission to the hospital for at least 24 consecutive hours.

Disease severity was defined as admission to ICU and/or death during hospitalization. We also estimated the length of hospital stay among the COVID-19 patients who were hospitalized and successfully discharged. Those who died during hospitalization were excluded from the length of hospital stay analysis.

Descriptive analyses were performed to summarize the characteristics of the COVID-19 patients overall and by hospitalization or disease severity status. Chi-square or Fisher's exact (for variables with a category including less than 5 participants) tests were used to estimate P-values for differences. The differences in the length of hospital stay were evaluated by means of Wilcoxon Rank Sum tests. Logistic regression was used to estimate the covariates-adjusted odds ratios (OR) and 95% confidence intervals (CI) for the associations of the patients' characteristics and comorbidities with hospitalization and disease severity. We also performed multinomial All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) 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 June 26, 2020. . https://doi.org/10.1101/2020.06.25.20137323 doi: medRxiv preprint logistic regression to further stratify disease severity into admission to ICU and death. To identify the factors associated with length of hospital stay, we fitted generalized linear models with a gamma distribution and a log link function which is similar in shape to the log-normal distribution and is robust to the outcome of length of hospital stay that was right skewed. 22 The diagnostic plots for the generalized linear models showed that the assumptions were reasonably met and the regression coefficients (β) along with the 95% CI for the associations were reported.

All the models were adjusted for age, sex, race/ethnicity, and cigarette smoking. The analyses were performed in SAS Version 9.4 (SAS Institute, Cary, NC) and two-sided p-values <0.05

were considered statistically significant in all analyses.

The 689 COVID-19 patients included in our study had a median age of 49.5 years (IQR: 35.2, 67.5); 53% were male. The race/ethnicity of study patients was 29.2% non-Hispanic White, 25 .5% non-Hispanic Black, 32.5% Hispanic, and 12.8% patients of 'Other' race/ethnicity. Cardiovascular disease was the most common comorbidity (49.5%). The other comorbidities included diabetes (24.7%), neoplasm or history of neoplasm (19.7%), obesity (18.6%), osteoarthritis (14.2%), vitamin D deficiency (12.9%), chronic kidney disease (11.8%), asthma (10.2%), COPD (8.8%), and pure hypercholesterolemia (2.9%). Hematological disorders were reported in 25.5% of patients for anemia, 8.0% for coagulation defect, and 5.4% thrombocytopenia ( Table 1) . All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) 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 June 26, 2020. . https://doi.org/10.1101/2020.06.25.20137323 doi: medRxiv preprint Two-hundred-sixteen (31.3%) of COVID-19 patients were hospitalized and 91 (13.2%) had severe COVID-19. The COVID-19 patients who were hospitalized and/or had severe disease tended to be older, to be non-Hispanic Black, and to be past or current smokers, compared to those who were not hospitalized or free of severe disease. Hospitalized patients and those with severe disease also tended to have comorbidities such as obesity, diabetes, pure hypercholesterolemia, asthma, COPD, chronic kidney disease, cardiovascular disease, neoplasm or history of neoplasm, osteoarthritis, vitamin D deficiency, and hematological disorders (anemia, coagulation defect, and thrombocytopenia) ( Table 1) .

Among the COVID-19 patients who were hospitalized and survived (N=191), the median length of hospital stay was 6.91 days (IQR: 3.27, 11.56) ( Table 2 ). This duration was higher in adults aged 60 years or older compared with younger individuals, in non-Hispanic Whites and non-Hispanic Blacks, in patients with diabetes, asthma, COPD, chronic kidney disease, cardiovascular disease, neoplasm or history of neoplasm, vitamin D deficiency, and hematological disorders (anemia, coagulation defect, and thrombocythemia) ( Table 2) .

In adjusted analysis, the sociodemographic characteristics associated with hospitalization were age (OR: (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. 2.64, 13.77]) were also associated with higher prevalence of hospitalization (Table 3) .

The sociodemographic characteristics associated with severe COVID-19, after adjustment for covariates were age had a strong positive association with severe COVID-19 (Table 3) .

In multinomial analysis stratifying COVID-19 severity into admission to ICU and death, being non-Hispanic Black compared to being non-Hispanic White was the only sociodemographic characteristic associated with both admission to ICU (OR: 3.32, 95% CI: 1.56, All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (Table 4 ).

In adjusted analysis, male sex (β: 0.39, 95% CI: 0.16, 0.62) was longer length of hospital stay while Hispanics (β: -0.40, 95% CI: -0.74, -0.06) and participants of 'Other' race/ethnicity (β: -0.65, 95% CI: -1.16, -0.14) had shorted length of hospital stay than non-Hispanic Whites. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. were also associated with prolonged length of hospital stay. (Table 5) .

In this racially and ethnically diverse study population of COVID-19 patients, the sociodemographic characteristics associated with hospitalization and/or disease severity included older age, non-Hispanic Black or Hispanic race/ethnicity (compared to non-Hispanic Whites).

Among the studied comorbidities, diabetes, pure hypercholesterolemia, asthma, COPD, chronic kidney disease, cardiovascular disease, osteoarthritis, and vitamin D deficiency were risk factors associated with hospitalization and/or severe COVID-19. Hematological disorders such as anemia, coagulation defects, and thrombocytopenia were associated with increased odds of both hospitalization and disease severity.

This report is one of a few epidemiological studies investigating factors for disease severity that was conducted in a racially and ethnically diverse sample of COVID-19 patients. In the UK, three studies analyzed data from the national Biobank and, consistent with our findings on non-Hispanic Blacks, found excess hospitalizations in Blacks and Asians compared to Whites. [23] [24] [25] In the US, five studies conducted among hospitalized COVID-19 patients observed an overrepresentation of non-Hispanic Blacks, with prevalences ranging from 51.0% to 83.2%. 9, 12, [25] [26] [27] However, the studies found that non-Hispanic Blacks were not at higher risk of severe disease or death compared to non-Hispanic Whites among hospitalized COVID-19

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. patients reported that non-Hispanic Blacks and Hispanics were more likely to test positive for COVID-19 than non-Hispanic Whites, but were not at higher risk of 30-day mortality.

Of all of these studies, only one examined length of hospital stay as an outcome and it found no difference in length of hospital stay by race or ethnicity. 12 Our results also suggested that although non-Hispanic Blacks and Hispanics were more likely to be hospitalized and to have severe COVID-19 than non-Hispanic Whites, length of hospital stay was not different between non-Hispanic Whites and non-Hispanic Blacks. It was even shorter in Hispanics and patients of 'Other' race/ethnicity than in non-Hispanic Whites. It is possible that the higher death rates in non-Hispanic Blacks could have affected the length of hospital stay and explained the non-significant difference with non-Hispanic whites. It is unclear why hospital stay was shorter in Hispanics and patients of 'Other' race/ethnicity.

Our results confirmed previous reports that older age and smoking as well as comorbidities such as diabetes, cardiovascular diseases, chronic kidney, COPD, coagulation defect, and thrombocytopenia are associated with hospitalization and/or severe COVID-19. 6, 9, 27, 29, 30 Consistent with the longer length of hospital stay noted in men in our analysis, it has been suggested that male sex may predispose to more severe COVID-19 due to hormonal factors. Although we found no association between obesity and severe COVID-19, other studies observed that morbid obesity (body mass index ≥40) was a risk factor of disease severity. 11, 27 With regards to patients with neoplasm or a history of neoplasm, they were not found to have higher COVID-19 severity in our study. Likewise, the evidence of an association between asthma and COVID-19 severity is also scarce and the previous studies were underpowered as they included rare asthma cases among COVID-19 patients. 6, 15 Consistent with our findings, a recent analysis of UK biobank data concluded that asthma was associated with severe COVID-19 among 641 infected patients and that this association was driven by non-allergic asthma. 17 Our epidemiological study is also the first to investigate the association of conditions such as vitamin D deficiency, pure hypercholesterolemia, osteoarthritis, and anemia with COVID-19 severity. An analysis of UK biobank data examined the serum levels of vitamin D and the risk of infection with COVID-19 and found no association, while another one performed among patients from the University of Chicago showed an increased risk of COVID-19 infection associated with vitamin D deficiency. 18, 35 However, the studies did not investigate the association of vitamin D and disease severity among patients who were already infected with SARS-CoV-2. 18, 35 In an ecological study, mortality rates among COVID-19 patients were compared among developed nations and it was suggested that countries with higher prevalence All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) 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 June 26, 2020. . https://doi.org/10.1101/2020.06.25.20137323 doi: medRxiv preprint 1 4 of vitamin D deficiency had higher mortality rates. 36 Yet, ecological studies do not include patient-level data and suffer from ecological fallacy, which is the assumption that factors associated with the national disease rates are associated with disease in individual patients. 37 Nevertheless, there is reason to investigate Vitamin D status as a factor in disease progression.

Vitamin D is well-known for its immunoregulatory properties; it can increase cellular immunity by stimulating antimicrobial peptides and could oppose cytokine storms induced by the innate immune responses. 38, 39 No published epidemiological study has also investigated whether high cholesterol is associated with higher risk of COVID-19 severity. Our study examined pure hypercholesterolemia indicative of familial hypercholesterolemia, which is a major contributor to atherosclerosis and cardiovascular disease, which itself, is associated with severe COVID- 19. 40 Furthermore, high cholesterol might enhance the replication of SARS-CoV-2 in endothelial cells potentially causing acute vascular injury and triggering coagulopathies. 41 No study has also examined the relationship of osteoarthritis and anemia with severe COVID-19. Osteoarthritis is a common joint disease, particularly prevalent in older adults; it is associated with various comorbidities and with an increased risk of mortality from cardiovascular diseases, diabetes and renal diseases. 42,43 Anemia affects 5.6% of all Americans, but this prevalence is increased to more than 10% in Americans aged 65 or older. 44,45 It is well-known to be a factor for poor prognosis in respiratory diseases such as COPD. 46 In our study however, it is unclear whether anemia is a factor for severe COVID-19 of or a consequence of the inflammation and cytokine production caused by the infection. 46 Our study has limitations. It included only patients from a single health system in the Midwest of the U.S. and may not be generalizable to the U.S. population overall. Our study design was observational; therefore, temporality and causality between certain factors (anemia, All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) 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 June 26, 2020. . https://doi.org/10.1101/2020.06.25.20137323 doi: medRxiv preprint coagulation defects, and thrombocytopenia for instance) and COVID-19 severity cannot be established. Smoking status was missing for 22.8% of the COVID-19 participants. Vitamin D deficiency was defined using electronic medical records and data on the serum vitamin D levels were not available. Nonetheless, our study has major strengths, it was conducted in a large and racially diverse sample of COVID-19 patients and included factors missing in previous reports such as vitamin D deficiency, pure hypercholesterolemia, osteoarthritis, and anemia.

In conclusion, the present study confirms factors previously reports that older age, non- (which was not certified by peer review) 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 June 26, 2020. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) 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 June 26, 2020. . https://doi.org/10.1101/2020.06.25.20137323 doi: medRxiv preprint All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) 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 June 26, 2020. . https://doi.org/10.1101/2020.06.25.20137323 doi: medRxiv preprint 1 Abbreviations: IQR, interquartile range COPD, chronic obstructive pulmonary disease; N, number of participants in subgroups. P-values for difference calculated using Wilcoxon rank tests.

All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) 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 June 26, 2020. (which was not certified by peer review) is the author/funder, who has granted medRxiv a license to display the preprint in perpetuity. (which was not certified by peer review) 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 June 26, 2020. . https://doi.org/10.1101/2020.06.25.20137323 doi: medRxiv preprint β , regression coefficient; CI, confidence interval; COPD, chronic obstructive pulmonary disease; Ref, reference. Regression coefficients calculated using generalized linear models with a gamma distribution and a log link function. Models adjusted for age, gender, race/ethnicity, and smoking. All rights reserved. No reuse allowed without permission.

(which was not certified by peer review) 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 June 26, 2020. . https://doi.org/10.1101/2020.06.25.20137323 doi: medRxiv preprint

The novel coronavirus originating in Wuhan, China: Challenges for global health governance

WHO Coronavirus Disease

SARS: Epidemiology, clinical presentation, management, and infection control measures

Clinical features of patients infected with 2019 novel coronavirus in Wuhan

Clinical characteristics of 138 hospitalized patients with 2019 novel coronavirus infected pneumonia in Wuhan, China

Risk factors for severity and mortality in adult COVID-19 inpatients in Wuhan

Clinical course and risk factors for mortality of adult inpatients with COVID-19 in Wuhan, China: A retrospective cohort study

Risk factors associated with acute respiratory distress syndrome and death in patients with coronavirus disease 2019 pneumonia in Wuhan, China

Predictors for severe COVID-19 infection

Predictors of mortality for patients with COVID-19 pneumonia caused by SARS-CoV-2: A prospective cohort study

Factors associated with hospital admission and critical illness among 5279 people with coronavirus disease

Older age and comorbidity are independent mortality predictors in a large cohort of 1305 COVID-19 patients in Michigan, United States

Asthma and COVID-19: Is asthma a risk factor for severe outcomes?

Clinical characteristics of 140 patients infected with SARS-CoV-2 in Wuhan

Asthma and the coronavirus disease 2019 pandemic: A literature review

Association of asthma and its genetic predisposition with the risk of severe COVID-19

Association of vitamin D deficiency and treatment with COVID-19 incidence

Total and high-density lipoprotein cholesterol in adults: United States

Vitamin D deficiency and insufficiency among US adults: Prevalence, predictors and clinical implications

Covid-19 by race and ethnicity: A national cohort study of 6 million united states veterans

Generalized linear models. Routledge

Race, socioeconomic deprivation, and hospitalization for COVID-19 in English participants of a national biobank

Ethnic disparities in hospitalization for COVID-19: A community-based cohort study in the UK

Ethnic and socioeconomic differences in SARS-CoV-2 infection: Prospective cohort study using UK biobank