Background
Despite ubiquitous progress in ART global scale up and significant reductions in mortality and morbidity some populations, such as fishing communities in Africa, continue to lag behind. We investigated the prevalence of, and factors associated with untreated HIV among fishermen using prospective data from a cluster randomized trial from Mangochi, Malawi.
Methods
This is analysis of cross-sectional data collected using in-person administered questionnaires between March 2022-January 2023 during the “creating demand for Fishermen’s Schistosomiasis and HIV services (FISH)” cluster randomised trial. The data came from 45 lakeside “boat-team” (clusters) in southern Lake Malawi from three trial arms. We defined the main outcome of having untreated HIV as a person living with HIV but currently not on ART using self-reported variables. Our main exposure was self-reported HIV testing history i.e. whether a person reported having ever tested for HIV before in their life. The analysis estimated the prevalence of untreated HIV and Binomial Exact 95% confidence intervals (CI) in the whole sample (N=5206) and among fishermen who tested HIV positive in the study (n=387). Unadjusted odds ratios (OR) and adjusted (aOR) were computed using random effects logistic regression while accounting for clustering.
Results
A total of 5206 fishermen were enrolled from 45 clusters: mean age 33.2 years (sd: 12.3), 3401/5206 (65.3%) were married, only 3032/5206 (58.2%) could read or write, and 4213/5206 (80.9%) had ever tested for HIV before. Prevalence of untreated HIV was 113/5206 (2.2%, 95%CI: 1.8; 2.6%) in the whole sample and 113/387 (29.2%, 95% CI: 24.9; 33.9%) among those who tested HIV positive in the study.
Having a previous HIV test was significantly associated with reduced odds of having untreated HIV both in unadjusted OR 0.05 (95%CI: 0.01; 0.16) and aOR 0.07 (95%CI: 0.02; 0.25) when adjusted for age, marital status and HIV knowledge variables (Table 1). There was evidence of clustering with intracluster correlation coefficient value of 0.12.
Conclusions
Not testing for HIV and increasing age were strongly associated with having untreated HIV. The findings encourage sustained investments in interventions to expand coverage of HIV testing in underserved populations.
Table or Figure
