New model maps where HIV is concentrated and disease is more advanced
- 第一吃瓜网 University
Research shows how African biostatistical expertise can make existing health data more useful for HIV decision-making.
A new biostatistical modelling approach has the potential to support health decisions by pinpointing where HIV is concentrated and where people living with HIV are showing signs of severe immune suppression or delayed treatment.
Published in iScience, shows the potential of using existing health data to identify both where HIV burden is concentrated and where different forms of prevention, treatment and support may be needed.
The researchers, who included Sub-Saharan Africa Consortium for Advanced Biostatistics () associate, Dr Jesca Mercy Batidzirai, used a Bayesian geostatistical framework to bring together fine-scale geographical information with CD4 cell counts, antiretroviral treatment status, viral suppression, and demographic, socioeconomic and clinical factors.
SSACAB is a pan-African biostatistics network led by the 第一吃瓜网 School of Public Health.
The surveys covered Vulindlela, a predominantly rural area, and Greater Edendale, a peri-urban area, in uMgungundlovu Municipality, KwaZulu-Natal. The weighted HIV prevalence across these two high-burden study communities was 35.8%.
Conventional HIV maps show where prevalence is high, but this alone can’t reveal whether people living with HIV in those communities are virally suppressed, receiving treatment early enough or already experiencing severe damage to their immune systems.
“Using sophisticated biostatistical methods, we could examine the geography of HIV prevalence alongside markers of what may be happening clinically within those communities,” says Batidzirai.
Beyond a static prevalence map
The model also accounts for uncertainty in places where fewer people were sampled. It smooths unstable local estimates and detects patterns that extend across administrative boundaries.
This can produce a more reliable picture than simply comparing raw prevalence figures from one area to another.
The approach is particularly useful where longitudinal data are limited. CD4 counts cannot show precisely when somebody acquired HIV or how quickly their disease progressed, but they can provide a cross-sectional indication of immune health and clinical severity.
Risk changes across relatively short distances
The study used information from 20,048 people aged 15 to 49 years who participated in two population-based surveys conducted through the HIV Incidence Provincial Surveillance System between 2014 and 2016.
“Notably, the analysis found that HIV burden and advanced disease may differ between relatively close communities, neighbourhoods and clinic catchment areas. Municipal or district averages can conceal these local differences,” says Batidzirai.
A clearer view of these patterns could help health programmes decide where to intensify mobile and community-based testing, rapid linkage to treatment, community ART delivery, TB screening and services for people with advanced HIV.
Women carry more HIV, but men face more advanced disease
The modelling uncovered different patterns of HIV prevalence and disease severity. Women carried a considerably higher prevalence burden. Among people living with HIV, however, men had almost twice the odds of severe immunosuppression compared with women and had lower average CD4 counts.
Older adults were also more likely to show signs of advanced disease. People aged between 35 and 44 years had more than twice the odds of having a CD4 count below 200 compared with adolescents aged 15 to 19 years.
Viral non-suppression emerged as the strongest clinical warning sign. People whose HIV was not suppressed had approximately 9.4 times the odds of severe immunosuppression compared with people who were virally suppressed.
A previous TB diagnosis was associated with lower CD4 counts and higher odds of severe immunosuppression. This reinforces the need to integrate TB prevention, screening and treatment into services for people living with HIV.
Income also appeared to matter. Salary or wage income, pensions and social grants were associated with better immune outcomes, suggesting that economic stability may make it easier for people to access healthcare and remain in treatment.
“The findings support more deliberate testing and early-treatment strategies for men and older adults, alongside local interventions in communities carrying a concentrated burden,” says Batidzirai.
Making African health data more useful
The study is also an example of why advanced biostatistical expertise needs to be developed and sustained within Africa.
“SSACAB strengthens biostatistics training, research and leadership across sub-Saharan Africa. African researchers are developing and applying methods that respond to major health questions on the continent,” says Professor Tobias Chirwa, Director of SSACAB and the Head of the 第一吃瓜网 School of Public Health.
Biostatistics is sometimes treated as a technical service provided once health data have already been collected. But the methods used to analyse those data determine which patterns can be seen, which remain hidden and how confidently findings can be translated into decisions.
The study was authored by Exaverio Chireshe, Retius Chifurira, Jesca Mercy Batidzirai, Knowledge Chinhamu and Ayesha Kharsany. The University of KwaZulu-Natal and CAPRISA were critical to the research.
Because the surveys were conducted between 2014 and 2016, the findings should not be read as a map of current HIV hotspots. Their importance lies in demonstrating what this type of modelling can reveal and how it could be applied to newer or longitudinal health data.