The impact of HIV stigma and discrimination on HIV testing, antiretroviral treatment, and viral suppression in Africa: a pooled analysis of population-based surveys

Background:

UNAIDS' ending AIDS targets include ensuring that less than 10% of people with HIV, at risk of HIV, or affected by HIV experience HIV stigma and discrimination. We aimed to estimate the effect of HIV stigma and discrimination on the HIV care cascade in eastern, southern, central, and western Africa—regions with the largest HIV burden worldwide.

Methods:

We used population-based surveys in countries in eastern, southern, central and western Africa from 2000 onward identified through previous reviews and data catalogues. Surveys eligible for inclusion were screened and selected if they contained at least one measure of HIV stigma and discrimination, and at least one HIV outcome (ie, HIV testing, antiretroviral therapy [ART] use, or viral suppression). The individual-level data from all included surveys were pooled for our analyses. We assessed three community-level HIV stigma exposures (continuous): discriminatory attitudes, perceived stigma, and shame of association with people with HIV. Among people with HIV, we examined individual-level, past-year anticipated, or experienced stigma in health-care settings (binary). Using Poisson regression, we estimated adjusted prevalence ratios (aPRs) for the effect of each stigma measure on past-year HIV testing, ART use, and viral suppression. We used segmented regression with a median breakpoint to model non-linear effects of community-level stigma.

Findings:

We identified 124 surveys in 38 countries published between 2000 and 2023. Stigma towards people with HIV was high. 51% of participants held discriminatory attitudes, 71% perceived HIV stigma, and 43% reported shame of association. Among people with HIV, 34% had anticipated, and 8% experienced, health-care stigma. Where the prevalence of community-level discriminatory attitudes was equal to or above the median, people were less likely to have tested for HIV (aPR 0·88 per 10% increase, 95% CI 0·87–0·88), and people with HIV had lower ART use (0·96 per 10% increase, 0·94–0·97) and viral suppression (0·96, 0·95–0·98). Similarly, communities with higher (≥median) shame of association levels had lower testing (aPR 0·97 per 10% increase, 95% CI 0·96–0·98) and viral suppression (0·98, 0·96–1·00). People with HIV who anticipated or experienced health-care stigma were 7% (95% CI 4–10) less likely to be on ART and 10% (6–14) less likely to be virally suppressed, possibly reaching 14% (7–21) and 16% (9–23) when adjusted for selection bias.

Interpretation:

HIV stigma and discrimination have measurable, widespread, negative effects on the full cascade of care: HIV testing, treatment, and viral suppression. These effects are most pronounced in communities with the highest levels of stigma. Stigma reduction strategies should be prioritised to improve health outcomes and reduce health inequities.

Funding:

Canadian Institutes of Health Research, Canada Research Chairs Program, and Wellcome Trust.