HIV is associated with a higher rate of COVID-19 diagnosis but no adverse outcomes

Background: Most available data on COVID-19 among persons with HIV (PWH) focuses on hospitalized patients, while COVID-19 risk among PWH relative to the general population remains inconclusive. Methods: Using a retrospective comparative cohort analysis, we included all adults with established primary care service at UC San Diego who underwent testing for COVID-19 from March to July 2020. We dichotomized the cohort into two groups, PWH and non-PWH. We used bivariate analyses to compare group differences using a clinical hierarchical order, including COVID-19 diagnosis, hospitalization, intensive care unit (ICU) admission, intubation, and death. Logistic regression models for each hierarchical clinical outcome included pre-specified covariates (demographics, diabetes mellitus, and obesity) and additional covariates at a 0.20 threshold via backward model selection. Interactions between HIV status and other covariates were included at the 0.05 threshold. Additional covariates included insurance type, hypertension, cardiovascular disease, chronic kidney, lung, liver disease, rheumatologic disease, cancer history, ACE inhibitor or angiotensin II receptor blocker use, and active tobacco, alcohol or illicit drug use, and history of mental health disorder. Results: Of 235609 participants, 3609 were PWH and 232000 non-PWH. Of them, 22% of PWH and 6% of non-PWH were tested for COVID-19. The PWH group had a higher proportion of younger individuals (76% vs 57%), males (85% vs 39%), non-whites (42 vs 35%), and a history of mental illness (58 vs 29%) than the non-PWH group. Of those tested, 7% of PWH and 2% of non-PWH tested positive for COVID-19. The adjusted odds ratio of COVID-19 diagnosis for HIV vs non-HIV was 4.32 (95% CI 3.09-6.04). No significant differences were observed for PWH compared to non-PWH in the proportion of patients hospitalized (15% vs 15%), admitted to ICU (10% vs 7%), requiring inotropic support (3% vs 4%), or died (3% vs 3%) but PWH required mechanical ventilation more frequently than those non-PWH (8% vs. 3%). HIV was not a significant predictor of hospitalization, ICU admission, or mortality in any models (see table);however, the limited number of events decreased the statistical power. Conclusion: In our cohort, PWH were tested and diagnosed more frequently than those without HIV for COVID-19. PWH had an increased risk of becoming infected with COVID-19, even when adjusted for demographics and comorbidities. HIV status did not affect hospitalization, ICU admission, or mortality. (Figure Presented).