Importance of Appropriate Comparison Group for Assessing COVID-19 Outcomes of People Living with HIV

Research suggests studies should adjust their cohort comparison group to account for demographics most represented by people living with HIV.

Existing studies have shown that people living with HIV (PLWH) are at higher risk for adverse outcomes upon contracting COVID-19, including severe symptoms and mortality.

Recent research published in HIV Medicine sought to examine HIV affects the COVID-19 outcome of hospitalized patients.

The investigators conducted a multi-center, retrospective matched cohort study of COVID-19 hospital inpatients according to HIV status. They included six hospital trusts across England, four located in London. Included in the study were PLWH aged 18 and older, admitted as hospital inpatients from February 1, 2020 - March 31, 2020, with a positive PCR test.

The comparator cohort was selected from HIV-negative inpatients admitted in the same timeframe. HIV-negative patients were matched with PLWH according to test date ± 7 days, age ± 5 years, gender, and index of multiple deprivation decile ± 1. The preferred outcome for all inpatients was clinical improvement or hospital discharge by their 28th day.

The investigators began the study assuming HIV status would have no impact on the primary outcome, and that 20% of the cohort would die or not recover by day 28 and the other 80% would recover in a median duration of six days. Data simulations suggested a sample size of 50 PLWH and 100 HIV-negative comparators; researchers increased this ratio by up to 1:3, hoping more comparators would increase statistical power.

The investigators found that the cumulative hazard of patients reaching the primary outcome was 43% lower for PLWH than HIV-negative patients. However, this was ameliorated after adjusting for ethnicity, frailty, baseline hypoxemia, duration of symptoms prior to baseline, BMI, and comorbidities (hypertension, chronic cardiac disease, chronic lung disease, active malignancy, diabetes, and chronic renal disease). The adjusted results showed a 30% reduced hazard for clinical improvement or discharge.

The study reported “HIV status was not associated with difference in mortality rates,” although PLWH were more likely to have a longer hospital stay (10 versus 7.5 days) and require a mechanical ventilator (23.5% versus 17.1%). Among the five immunosuppressed PLWH, there were no deaths, although the average length of hospital stay was 11 days and 40% required a ventilator.

The investigators noted that the study was limited by its consideration of exclusively hospitalized people, as factors that led to hospitalization may affect outcomes.

Additionally, there were very few immunosuppressed PLWH in this study, so it would not be possible to apply these results of the severity of COVID-19 outcomes to persons with low CD4 counts.

They also reported a disproportionate impact of COVID-19 on Black, Asian, and minority ethnic persons, even after adjusting the results for age, sex, and comorbidities.

People living with HIV were frailer, though increased frailty was correlated with advanced age.

Because PLWH are more likely to be of Black, Asian, and minority ethnic backgrounds, more likely to be frail, and have a higher proportion of active malignancy, the investigators highlighted the importance of using an adjusted comparison group to assess outcomes of PLWH hospitalized with COVID-19.

The investigators recommended further study into the impact of the COVID-19 vaccinations and novel variants of COVID-19 in PLWH and immunosuppressed persons, but concluded by saying subsequent studies should adjust their results for confounding comorbidities and demographics to ensure an accurate comparison group.