HIV associated with worse outcomes from COVID-19 in UK ISARIC and OpenSAFELY databases
Simon Collins, HIV i-Base
Two UK studies, both published online ahead of peer review in the same week have reported worse outcomes from COVID-19 in HIV positive people compared to HIV negative general population. [1, 4]
Both have limited data on HIV history including details on ART, CD4 count and viral load, which for HIV-related research are essential.
Also, while one of the studies has been included as evidence for the UK Scientific Advisory Group for Emergencies (SAGE) the other prompted a statement from BHIVA and several community groups that cautioned the results ahead of peer review. [3, 6]
UK ISARIC database in people hospitalised with COVID-19
A UK cohort study based on one of the national COVID-19 databases has reported worse outcomes for HIV positive people compared to HIV negative people in the general population hospitalised with COVID-19. 
This was in an analysis that adjusted for baseline demographics, HIV and COVID-19 factors (age, gender, ethnicity), a series of 10 co-morbidities, whether the SARS-CoV-2 infection was acquired while already in hospital, and severity of disease at time of admission (to account for the fact that the decision to admit into hospital may be increased in someone known to have HIV).
The study was prompted by an early concern about the lack of data on HIV and COVID-19, including the lack of routine inclusion of HIV status on hospitalisation. Also, that early reports were generally case studies and very small cohorts. Funders included NIHR, the MRC, Wellcome and Public Health England.
The main ISARIC cohort was started on 17 January 2020 and data was included until cut-off of 18 June for this analysis, with primary outcome of cumulative mortality at day-28.  HIV researchers collaborated with the ISARIC group from March 2020.
The analysis included records from 47,539 patients who were hospitalised with laboratory confirmed (or highly likely) COVID-19 from 207 centres across the UK enrolled into the ISARIC CCP-UK study. Of these, 115 (0.24%) were confirmed HIV positive and 103/115 (89%) had a record of antiretroviral therapy.
At baseline, HIV positive people were significantly younger [median age 55 (IQR: 49 to 61) vs 74 (IQR: 60 to 84) years; p<0.001), had a higher prevalence of obesity (18% vs 11%, p=0.03), and moderate/severe liver disease (5.4% vs 1.9%, p=0.008), higher lymphocyte counts (p<0.001) and C-reactive protein (p=0.02), and more systemic symptoms.
Other difference at baseline included black ethnicity (45% vs 3%), fewer comorbidities [median 1 (IQR: 0 to 2) vs 2 (IQR: 1 to 3)], and less cardiovascular disease (18% vs 32%) and dementia (2.7 to 16.8), all p<0.001)
Although cumulative mortality was non-significantly lower in the HIV group (25.2% vs 32.1%, p=0.12) in the initial analysis that did not take into consideration confounding factors, stratification for age revealed a higher mortality among HIV positive people aged below 60 years. There was no significant effect of either gender or ethnicity, and taking into account co-morbidities did not modify the estimates. After considering the severity of disease at presentation (as indicated by the needs for oxygen therapy), HIV positive people showed a 63% increased risk of mortality compared to the HIV negative group (aHR 1.63; 95%CI: 1.07 to 2.48; p=0.02).
In the HIV positive group, mortality was more common among who presented with more advanced disease (hypoxia: 75% vs 38%, p=0.001) with higher heart rate (p<0.001 and among people with obesity and diabetes with complications).
However, important factors about HIV history, including nadir and current CD4 cell count, viral load and history of complication were not routinely collected in the study. Use of ART was included but often without details. The study is currently connecting ISARIC data to HIV clinic records.
The study has been accepted as evidence by the UK Scientific Advisory Group for Emergencies. 
The study group also conducted a data review of HIV and COVID-19 coinfection studies that is currently in press
UK OpenSAFELY primary care database
A second UK study, also published online ahead of peer review in the same week as the ISARIC study, also reported worse outcomes from COVID-19 in HIV positive people. 
This used the OpenSAFELY database that uses primary care medical records from GP surgeries linked to national death registrations in the UK from 1 February to 22 June 2020. Although this study is introduced as being “on behalf of the NHS” the research was not directly funded and OpenSAFELY was set up in response to COVID-19 by the medical journalist Ben Goldacre, working with NHS and other researchers.
OpenSAFELY contains the medical records of 17.3 million adults (approximately one-third of the UK population) of which 27,480 (0.16%) include a record of HIV positive status. The most significant baseline differences between the two groups included higher percentage of men (65% vs 50%), Black ethnicity (26% vs 1.9%), more social deprivation (ie 31% vs 19% in the most deprived category) and more chronic liver disease (3.4% vs 0.6%) in HIV positive vs negative groups respectively. By contrast, other factors, including median age [48 years (IQR: 40 to 55) vs 49 (IQR: 34 to 64)], BMI, smoking status and other comorbidities (diabetes, hypertension, asthma, heart disease, COPD, kidney disease and cancer etc) were broadly similar.
There were 25 (0.087%) vs 14,857 (0.038%) deaths in people with COVID-19 in the HIV positive vs negative groups respectively. After adjusting for age and sex, HIV was associated with a nearly 3-fold higher risk of death (HR 2.90, 95% CI: 1.96 to 4.30). This was reduced after further adjustment for deprivation, ethnicity, obesity, smoking, and comorbidities (HR 2.30, 95%CI: 1.55 to 3.41).
Black ethnicity had a greater association between HIV and COVID-19 death (HR 3.80, 95%CI: 2.15 to 6.74) compared with other ethnic groups (HR 1.64, 95%CI: 0.92 to 2.90), p-interaction=0.045.
Perhaps most surprisingly, and in contrast to other COVID-19 research, comorbidities did not have a strong association with poorer outcomes.
Although this study was based on a very large database the limitations of primary care records include under-reporting of HIV status and lack of any HIV-related history including ART status and CD4 and viral load history.
The ISARIC study is the largest UK cohort to date on HIV and COVID-19 coinfection and OpenSAFELY is the largest general population database to look at HIV.
Both studies show the importance of recording HIV status for all people hospitalised with COVID-19 and being able to adjust for demographic, HIV and COVID-19-related factors.
They still show the need for further research, which in a setting with electronic medical records should be readily possible.
A large study from South Africa, also reported an approximately 2-fold increased risk of mortality compared to the general population. Although it didn’t adjust for TB, COVID-19 outcomes were not significantly different for people with viral suppression on ART. 
Similar to the South African study, these UK papers report that being HIV positive might increase the risk of mortality with COVID-19 compared to the general population. This is independent of gender or ethnicity and comorbidities including obesity and diabetes and the ISARIC results were especially in younger people (less than 60).
A joint statement from BHIVA and several HIV organisations (THT, NAT and NAM) highlighted additional concerns with the OpenSAFELY study. This included limitations with coding and HIV records in primary care records, missing data on weight and incorrect attribution of COVID-19 as a cause of death. 
BHIVA is also planning to collect data from all HIV clinics to investigate the risk of worse COVID-19 outcomes in people living with HIV.
Simon Collins is a community representative with the UK group and is a coauthor on the paper.
- Geretti AM et al. Outcomes of COVID-19 related hospitalisation among people with HIV in the ISARIC WHO Clinical Characterisation Protocol UK Protocol: prospective observational study. Pre peer review. doi: 10.1101/2020.08.07.20170449. (11 August 2020).
- International Severe Acute Respiratory and Emerging Infection Cohort (ISARIC) clinical characterisation protocol (CCP).
https://isaric.tghn.org/ (main website)
- UK SAGE, papers for August 2020 meeting.
- Bhaskaran K et al. HIV infection and COVID-19 death: population-based cohort analysis of UK primary care data and linked national death registrations within the OpenSAFELY platform. Pre peer review. DOI: 10.1101/2020.08.07.20169490v1. (7 August 2020).
- Clayden P. HIV positive people in South Africa at increased risk of dying from COVID-19: first data from country with high prevalence of HIV and TB. HTB (26 June 2020).
- BHIVA and others. HIV and COVID-19: a statement on the Open SAFELY pre-print. (13 August 2020).
This article was originally posted on 12 August 2020.