COVID-19 Outcomes Among Persons Living With or Without Diagnosed HIV Infection in New York State

Key Points Question Is there an association between prior diagnosis of HIV infection and coronavirus disease 2019 (COVID-19) diagnosis, hospitalization, and in-hospital death among residents of New York State? Findings In a cohort study of linked statewide HIV diagnosis, COVID-19 laboratory diagnosis, and hospitalization databases, persons living with an HIV diagnosis were more likely to receive a diagnosis of, be hospitalized with, and die in-hospital with COVID-19 compared with those not living with an HIV diagnosis. After demographic adjustment, COVID-19 hospitalization remained significantly elevated for individuals with an HIV diagnosis and was associated with elevated mortality. Meaning Persons living with an HIV diagnosis experienced poorer COVID-related outcomes (principally, higher rates of severe disease requiring hospitalization) relative to those without an HIV diagnosis.


Introduction
Coronavirus disease 2019 (COVID-19) has resulted in more than 1.6 million deaths worldwide as of December 15, 2020, with the United States reporting the most diagnosed cases (n = 16 520 408) and deaths (n = 300 494). 1 In addition to having an older, more male demographic distribution, persons living with diagnosed HIV infection have a higher prevalence of many underlying medical conditions associated with more severe COVID-19 illness. [2][3][4] The Centers for Disease Control and Prevention identifies older adults and those with certain underlying medical conditions as being at elevated risk for severe illness from COVID- 19 were older than 50 years, and 72% were non-Hispanic Black (40%) or Hispanic persons (32%). 18 A recent match of surveillance databases within NYC compared the features and outcomes of persons living with diagnosed HIV with those of persons living with diagnosed HIV with COVID-19, but it did not estimate outcome rates or adjust for confounding factors associated with both infections. 16 We conducted a population-level match of NYS's HIV registry against its COVID-19 diagnosis and hospitalization databases, to provide a full population-level comparison of COVID-19 outcomes between the persons living with diagnosed HIV and persons living with diagnosed HIV within a US jurisdiction. Specifically, we compared the continuum of rates of COVID-19 diagnoses, hospitalizations, and in-hospital deaths for persons living with diagnosed HIV in NYS with those for persons living without diagnosed HIV and assessed factors associated with these outcomes among persons living with diagnosed HIV. 19

Study Population and Data Sources
This is a retrospective cohort study of individuals with polymerase chain reaction-confirmed severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection diagnosed between March 1 and June 7, 2020, in NYS. Data were abstracted from the following: (1)

Study Variables
The demographic variables included in all analyses were age, sex, and region of residence (Long Island, Mid-Hudson, NYC, and rest of NYS  [25][26][27] Statistical Analysis COVID-19 diagnoses, hospitalizations, and in-hospital deaths were evaluated between persons living with diagnosed HIV and persons without diagnosed HIV. 28 Proportion, or attack rates, expressed per 1000 persons, as well as attack rates per previous outcome (hospitalized per diagnosis and in-hospital death per hospitalization) and unadjusted rate ratios (RRs) with 95% CIs were calculated to evaluate associations between study variables and each outcome. Adjusted comparisons were made via indirect standardization, controlling for age, sex, and region. 29 Among persons living with diagnosed HIV, the attack rates per 1000 persons and per previous outcome, with unadjusted RRs with 95% CIs, were assessed among levels of age, sex, region, race/ ethnicity, transmission risk, care status, HIV stage, and viral suppression status. Adjusted RRs (aRRs) with 95% CIs were calculated in multivariable Poisson regression models, which included covariates significant at α = .05 in bivariate analyses. Viral suppression was not included in multivariable models because it is a potential mediator of the association between HIV stage and COVID-19 outcomes. All analyses were conducted using SAS, version 9.4 (SAS Institute Inc). 22       or 14% to 25% of total lymphocytes (aged Ն6 years) or CD4 of 500 to 999 cells/mm 3 or 22% to 29% of total lymphocytes (age 1-5 years); and stage 3 is CD4 less than 200 cells/mm 3 or less than 14% of total lymphocytes (aged Ն6 years) and CD4 less than 500 cells/mm 3 or less than 22% of total lymphocytes (aged 1-5 years). 25 Persons classified as "other" include those out of care for more than 36 months, who have moved out of state, and who have not moved but are receiving care out of state.

Discussion
To our knowledge, our study represents the first population-level match of an entire US state's HIV registry against its COVID-19 diagnosis and hospitalization databases, establishing state-level rates of COVID-19 outcomes among persons living with diagnosed HIV and comparisons with those observed in the overall population.

COVID-19 Hospitalization and Mortality
We found that persons living with diagnosed HIV were significantly more likely than persons living

Conclusions
Although the mechanisms underpinning increased risk are not fully understood, the intersection of HIV and COVID-19 has multiple implications. Because HIV infection is a marker for, and may play a direct role in, more severe COVID-19 outcomes, persons living with diagnosed HIV (with any CD4 count) may warrant recategorization from "might be at increased risk" to "increased risk" in the Centers for Disease Control and Prevention's underlying medical conditions list. 5 This change may lead to higher prioritization of persons living with diagnosed HIV for receipt of the COVID-19 vaccine, per national and state allocation plans. 48,49 Finally, a syndemic association between these infections may act multiplicatively on affected persons and communities, which are more likely to involve persons of color and urban areas. 50 Our findings present an opportunity to address health equity with regard to HIV and COVID-19 through a combination of prevention and treatment approaches. 51