RESUMO
OBJECTIVES: The aim of this study was to identify community testing modalities associated with fast-track ART initiation in Botswana. METHODS: We conducted a retrospective cohort study that included all Botswana citizens 15 years or older who were newly identified as HIV-positive from 1 May 2017 to 31 January 2019, in Mahalapye and Southern districts. We used Poisson regression with robust error variance and generalised linear mixed models to control for cluster effects to model risk of ART initiation within 7 and 30 days of HIV diagnosis, testing modality factors. RESULTS: A total of 1436 individuals were newly identified HIV-positive, with men accounting for 60% across all testing modalities. 22% of all HIV-positive individuals were initiated on ART within 7 days. Clients diagnosed through index testing were more likely to be started on ART within 7 days (adjusted risk ratio [aRR] = 1.38, 95% CI 1.37-1.38) and 30 days (aRR = 1.17, 95% CI 1.09-1.26) than those diagnosed through mobile/outreach testing. CONCLUSIONS: Community HIV testing can complement facility-based testing by reaching individuals who may be less likely to seek HIV services at a facility, such as men. Monitoring ART initiation by testing modalities is critical to identify the optimal ones and to guide continuous programme improvement.