The risk of severe COVID-19 and mortality from COVID-19 in people living with HIV compared to individuals without HIV - a systematic review and meta-analysis of 1 268 676 individuals

  1. Lovemore Mapahla
  2. Asmaa Abdelmaksoud
  3. Rida Arif
  4. Nazmul Islam
  5. Albert Chinhenzva
  6. Suhail A. R. Doi
  7. Tawanda Chivese

Reviewed by

Read the full article

Listed in

Log in to save this article

Abstract

Background

There is conflicting evidence about the risk of mortality and severe disease due to COVID-19 in people living with HIV (PLHIV).

Objectives

To compare mortality, hospitalization, and the need for intensive care services due to COVID-19 between PLHIV and individuals without HIV based on data from the existing literature.

Methods

A comprehensive search in PubMed, Cochrane Library, Scopus, China Academic Journals Full Text Database, the Database of Abstracts of Reviews of Effectiveness (DARE) and and the medRXIV and bioRxiv databases of preprints was carried out. Each data source was searched from 1 January 2020 to 20 th of February 2021. Eligible studies were case control, cross-sectional and cohort studies where participants had confirmed COVID-19. From each study, data on numbers of PLHIV and individuals without HIV for each outcome were extracted. Study quality was assessed using the MethodologicAl STandard for Epidemiological Research (MASTER) scale. Data synthesis used a bias adjusted model and predefined age and geographical subgroups were analysed.

Results

Of a total of 2757 records identified, 11 studies, from 4 countries, the United Kingdom, Spain, the United States of America and South Africa, were included. The total participants assessed for the outcomes in this meta-analysis were 1 268 676 of which 13 886 were PLHIV. Overall, the estimated effect of HIV on mortality suggested some worsening (OR 1.3, 95% CI: 0.9 – 2.0, I 2 = 78.6%) with very weak evidence against the model hypothesis at this sample size. However, in individuals aged <60 years, the estimated effect on mortality suggested more worsening in PLHIV (OR 2.7, 95% CI: 1.1 - 6.5, I 2 = 95.7%) with strong evidence against the model hypothesis at this sample size. HIV was also associated with an estimated effect on hospitalization for COVID-19 that suggested worsening (OR 1.6, 95% CI: 1.3-2.1, I 2 = 96.0%) also with strong evidence against the model hypothesis at this sample size. A secondary analysis of the included studies suggested no difference, by HIV status, in the prevalence of pre-existing conditions.

Conclusion

People living with HIV have higher risk of death and hospitalisation from COVID-19, compared to individuals without HIV. A secondary analysis suggests this is not due to associated comorbid conditions. The difference in mortality is exaggerated in those younger than 60 years of age.

Registration

PROSPERO: CRD42020221311 ( https://www.crd.york.ac.uk/prospero/display_record.php?RecordID=221311 )

Evidence before this study

Findings from existing studies have shown conflicting evidence concerning the risk of severe COVID-19 and death from COVID-19 in people living with HIV (PLHIV) compared to people without HIV. Evidence from three existing systematic reviews suggests that the risk of severe COVID-19 and death from COVID-19 in PLHIV may be similar to that in individuals without HIV. However, findings from three large cohort studies and one meta-analysis of four studies suggest that the risk of death from COVID-19 in PLHIV may be higher than that in individuals without HIV. One of the large cohort studies, which is also included in the previous meta-analysis, consisted of individuals with unknown COVID-19 status, and therefore there is still debate concerning the risk of severe COVID-19 outcomes in PLHIV.

Added value of this study

In this meta-analysis of 11 studies with 1 268 676 individuals with confirmed COVID-19, we found a stronger difference in mortality by HIV status for those individuals below the age of 60 years, and over this age, HIV had an attenuated effect on mortality, suggesting that age-related mortality overshadows PLHIV related mortality. Further, PLHIV had increased odds of being hospitalized and needing intensive cares services, probably related to increased COVID-19 severity in PLHIV. A secondary analysis of the included studies suggested no difference in the prevalence of pre-existing conditions.

Implications of all the available evidence

Our findings suggest that PLHIV are at higher risk than the general population and should be prioritized for vaccine coverage and monitoring if diagnosed with COVID-19. This is especially important for countries in Sub-Saharan Africa that have a high burden of HIV in the younger populations who are more vulnerable.

Strengths

This study was carried out rigorously following the PRISMA guidelines for systematic reviews and meta-analyses. We used a comprehensive search strategy across most of the main citation databases to ensure that no relevant studies were missed. We included studies where participants had confirmed COVID-19 only and we synthesized the findings from studies using a bias adjustment model that took into consideration the quality of included studies.

Limitations

All studies included in this review are observational studies and conclusions about causality require cautious interpretation. Due to a lack of data from included studies, we were not able to analyse the effect of being on treatment for HIV, and HIV control variables such as viral load and CD4 counts on COVID-19 hospitalization, intensive care services and mortality. Lastly most of the included studies had small samples overall or for PLHIV and this may affect the effect estimates in this analysis. Future research is therefore indicated to confirm these findings.

Article activity feed

  1. ScreenIT

    SciScore for 10.1101/2021.07.03.21259958: (What is this?)

    Please note, not all rigor criteria are appropriate for all manuscripts.

    Table 1: Rigor

    Ethicsnot detected.
    Sex as a biological variablenot detected.
    RandomizationTh assumption of this quality effects model is that there is one true effect size shared by all included studies and any deviations from this due to random and/or systematic error.
    BlindingSix reviewers blindly screened titles and abstracts of studies uploaded on Rayyan platform Rayyan (http://rayyan.qcri.org/) (22) and conflicts were resolved through discussion.
    Power Analysisnot detected.

    Table 2: Resources

    Software and Algorithms
    SentencesResources
    Search strategy and data sources: We searched for studies in the following electronic databases; PubMed, Cochrane Library, Scopus, China Academic Journals Full Text Database, the Database of Abstracts of Reviews of Effectiveness (DARE) and and the medRXIV and bioRxiv databases of preprints.
    Cochrane Library
    suggested: (Cochrane Library, RRID:SCR_013000)
    bioRxiv
    suggested: (bioRxiv, RRID:SCR_003933)
    We also carried out a citation search of the top 20 similar articles of all included studies on PubMed to retrieve studies that might have been missed in the original electronic search.
    PubMed
    suggested: (PubMed, RRID:SCR_004846)

    Results from OddPub: We did not detect open data. We also did not detect open code. Researchers are encouraged to share open data when possible (see Nature blog).

    Results from LimitationRecognizer: We detected the following sentences addressing limitations in the study:
    Our study has several limitations. The studies included in this synthesis are observational studies and any inferences about the strength of the relationship between HIV and severe COVID-19 need to be made cautiously. Further, confounding variables, such as age and comorbidities, and type of HIV treatment may have affected some of the findings from included studies. It was not possible to do a meta-analysis of adjusted effect sizes as the studies reported different effect measures. We were also not able to carry out subgroup analysis on some of these possible confounders because of lack of data from included studies. Due to a lack of data from included studies, we were not able to analyse the effect of being on treatment for HIV, viral load and CD4 counts on COVID-19 severity and mortality. However, it is worth noting that one study found that the type of anti-retroviral treatment had no beneficial impact on COVID-19 severity or risk of death (13). A strength of this study is that it was carried out rigorously using the updated PRISMA guidelines for systematic reviews and meta-analyses. Further, this meta-analysis included studies where participants had confirmed COVID-19 only and used quality adjusted effects model to take into consideration the study quality of included studies, and therefore the risk estimates we reported are adjusted for the variations in study quality of the included observational studies.

    Results from TrialIdentifier: No clinical trial numbers were referenced.

    Results from Barzooka: We did not find any issues relating to the usage of bar graphs.

    Results from JetFighter: We did not find any issues relating to colormaps.

    Results from rtransparent:
    • Thank you for including a conflict of interest statement. Authors are encouraged to include this statement when submitting to a journal.
    • Thank you for including a funding statement. Authors are encouraged to include this statement when submitting to a journal.
    • No protocol registration statement was detected.

    Results from scite Reference Check: We found no unreliable references.

    About SciScore

    SciScore is an automated tool that is designed to assist expert reviewers by finding and presenting formulaic information scattered throughout a paper in a standard, easy to digest format. SciScore checks for the presence and correctness of RRIDs (research resource identifiers), and for rigor criteria such as sex and investigator blinding. For details on the theoretical underpinning of rigor criteria and the tools shown here, including references cited, please follow this link.

  2. Version published to 10.1101/2021.07.03.21259958v2 on medRxiv
  3. Version published to 10.1101/2021.07.03.21259958v1 on medRxiv