Study population and setting
The study described population risk factors for COVID-19 infection, hospitalization, and death among adults living in non-congregate settings in Denver, CO from March 6 to October 6, 2020. The Denver Health and Hospital Authority was notified of laboratory-confirmed COVID-19 cases in the city and county of Denver via notification from the Colorado Electronic Disease Reporting System. They conducted interviews with case-patients and reviewed their medical records. They gathered information on age, gender, race and ethnicity, potential sources of exposure, signs and symptoms, and if they continued to work while they were experiencing symptoms. Race/ethnicity was defined as non-Hispanic white, non-Hispanic Black, Hispanic, Asian, American Indian/Alaskan Native, Native Hawaiian/Pacific Islander, other races/ethnicities not previously specified (such as Burmese, Egyptian, or Filipino) or mixed race. For cases reported after May, additional data were collected on occupation, industry of occupation, and household size, and were therefore available for a subset of cases.
Summary of Main Findings
The study identified 10,163 adults diagnosed with SARS-CoV-2 infection in non-congregate settings living in Denver city or county. 1,087 people were hospitalized (10.7%) at the time of their diagnosis, and 165 (1.6%) died during the 7-month study period. Race and ethnicity data were available for the majority of cases (9,056/10,163); among those with data, 54.8% (N=4,959) were Hispanic, and they also accounted for 51.2% of deaths (N=62). White patients made up 32.3% (N=2,926), while only making up 18.2% of hospitalized patients (N=167) and 29.8% of deaths (N=36). Black participants made up a small proportion of total cases (6.4%, N=579), but 11.5% (N=105) and 11.6% (N=14) of hospitalizations and deaths respectively. The study plotted cases over the study period as well, and there were 3 notable waves of transmission among Hispanic participants, peaking in April-May, in July-August, and in September-October, while for Black, American Indian/Alaskan Native, Asian, and Native Hawaiian/Pacific Islander cases, there were no pronounced waves. Finally, Hispanic cases were more likely to work in an essential industry, work while symptomatic, and have an exposure through a household contact, compared to non-Hispanic cases.
The study had data from all diagnosed cases in Denver city and county, which reduced selection bias to ensure generalizability to all individuals living in Denver. They also collected important epidemiological data on potential occupational and personal transmission routes that could potentially explain some of the racial/ethnic disparities seen in the sample.
The study was limited by passive ascertainment of cases through the notification system, which may have a delay in reporting or limited information for outcome data following the interview with Denver Public Health staff. Additionally, an unspecified number of patients may not have responded to interview requests due to being severely ill or having died before they could be contacted, which would introduce selection bias and deflate the number of cases included in this study. There were also information limitations on earlier interviews, as the occupation questions were only added in May, so results about occupation may not be applicable to cases from earlier in the pandemic. There also may have been differences in care-seeking for testing or availability of testing sites based on race/ethnicity, which could result in Hispanic or Black residents being diagnosed at the hospital or at death, rather than in community testing sites.
Many prior studies of disparities in risk factors for COVID-19 diagnoses, hospitalizations and deaths have focused on Black vs. white cases, but this study explores heath disparities for Hispanic people and potential occupational and familial or contact-based reasons why these may occur.
This review was posted on: 22 January 2021