MATERIALS AND METHODS
Data on COVID-19 test results were collected by NYC using syndromic and reportable disease surveillance strategies from February 29 (first confirmed case March 1, 2020) to June 12, 2020. After IRB approval, the NYC Department of Health and Mental Hygiene (NYCDOHMH) provided COVID-19 sex-specific age-matched data that included total number of tests, cases, hospitalizations, and deaths. Separately, data on the presence of unspecified comorbid conditions were collected through May 16, 2020 but were only available by gender and were not age-matched.6 Only individuals with first-time viral tests were included, and those with unknown sex or age data were excluded.
Exposure and Covariates
Exposure of interest was self-reported sex (male or female only; transgender and gender-nonconforming were excluded). Data were stratified into age categories (0–17, 18–44, 45–64, 65–74, and 75+ years) per the NYCDOHMH based on trend similarity within each subgroup.
Outcomes of interest were COVID-19 case positivity, hospitalization rate and death rate. Cases included any New York City resident tested and/or treated in NYC and were defined based on positive nasopharyngeal swabs for SARS-CoV-2. Hospitalizations included reports from hospitals, hospital system and Regional Health Information Organizations, and additional sources were used when data were incomplete. Deaths were a combination of confirmed (NYC resident with positive COVID-19 test) and probable (NYC resident, or pending residency, with no known positive COVID-19 test but death certificate lists COVID-19 or an equivalent as cause of death) and were obtained from the Health Department’s surveillance database.
Case positivity, hospitalization and death rates were tabulated utilizing total tests as the denominator and were sex- and agestratified. Rates per 100,000 individuals in NYC were further computed using 2018 United States Census Bureau data . Relative risks (RR) of positive cases, hospitalization and death were calculated and stratified by age and sex. Relative risks for comorbidities were only calculated by gender. Sensitivity analyses were completed for hospitalization and death utilizing positive cases as the denominator. Case fatality rates (CFR) were calculated as percentages using deaths (probable and confirmed) as a proportion of the total number of positive cases. Missing data were not included, and therefore only complete case analysis was performed. Statistical significance was evaluated at an alpha of 0.05 and model estimates are presented with 95% confidence intervals. All analyses were performed using Stata v14.
When stratified by age (Figures 1 and 2, Table 1), in all categories except those aged 0-17, the majority of tests were completed in women. With respect to cases, the majority occurred in men (except those >75 years), with the most significant sex difference observed in the 65–74 age group. Hospitalizations demonstrated a similar pattern with the greatest difference observed in the 45–64 age cohort (62.63%, n=11,189). The likelihood of COVID-19 death, in every age category, was higher in men, with the greatest differences in 18–44 and 45–64 age groups (74.35%, n=600 and 69.01%, n=3365, respectively).
Using gender-stratified population level data (per 100,000), overall rates of testing were almost identical in men and women, and all three outcomes of interest were greater in men (Table 1). Further stratification by age revealed that every age category displayed greater rates of case positivity, hospitalization and death for men compared to women.
The relative risk of case positivity, hospitalization and death was significantly higher in men when compared to women (RR 1.59, 95%CI 1.55–1.64 for death, Table 1). This difference persisted after age-stratification (Table 1). The greatest difference in case positivity was observed in men aged 65–74 years (RR 1.22, 95%CI 1.19–1.24). The greatest difference in hospitalization was observed in men aged 45–64 years (RR 1.85, 95% 1.80–1.90), while the greatest differences in death were in men aged 18– 44 years and 45–64 (RR 3.30, 95% CI 2.82–3.87, and RR 2.46, 95%CI 2.31–2.61, respectively). For all remaining age-stratified categories, except for those aged 0-17, and for each outcome, men demonstrated a statistically significant increased risk compared to the age-matched group of women.
Hospitalizations and deaths due to COVID-19 in men and women were also compared to the total number of positive tests as opposed to the total number of tests (Table 2). Similarly, almost all age-matched comparisons demonstrated greater risks for men compared to women.
Analysis of comorbidity data (Table 3) revealed 78.5% of men who endorsed a comorbidity in comparison to 80.1% of women.