HIV raises TB prevalence in Southern and Eastern Africa

Tuberculosis (TB) and human immunodeficiency virus (HIV) remain two of the world’s deadliest infectious diseases, especially in low- and middle-income countries (LMICs). Despite known biological links between the two infections, there has been surprisingly little systematic evidence on how the number of people living with TB in the community differs by HIV status. Corresponding author Nicole A. Swartwood and colleagues set out to fill that gap by conducting a broad systematic review and meta-analysis of TB prevalence surveys conducted in LMICs between January 1, 1993 and October 13, 2025 (PROSPERO CRD42024503853). The team screened 10,211 potentially relevant publications and extracted community-representative, bacteriologically-confirmed TB prevalence estimates that were stratified by participant HIV status. Their primary analysis focused on surveys that systematically offered HIV testing to participants who were eligible to give sputum or who were TB-positive, ensuring the HIV status data were not just self-reported. While the full collated dataset spanned 681,402 participants across ten countries, the core pooled analysis used 12 surveys representing 264,530 participants within nine countries in Southern and Eastern Africa, providing the clearest window into how TB prevalence compares between people living with and without HIV.

To compare TB prevalence by HIV status, the researchers included surveys that reported bacteriologically-confirmed TB and that had systematic HIV testing protocols for participants. They extracted prevalence numbers for people living with versus without HIV and used Bayesian meta-regression to combine results across studies and estimate pooled risk ratios (RR). In the primary analysis, 11 of 12 surveys reported higher TB prevalence among participants living with HIV than among those without. The pooled RR was 3·86 (95% credible interval: 2·41–5·53), indicating roughly a three- to six-fold higher prevalence of bacteriologically-confirmed TB among people living with HIV in these settings. The team also compared routine notification data to survey prevalence by calculating country-level and overall TB notification-to-prevalence (N:P) ratios by HIV status. N:P ratios were higher among participants living with HIV in all examined countries; the overall pooled N:P ratios were 1·74 (0·59–4·56) for people living with HIV and 0·48 (0·17–1·20) for people without HIV. The study was registered in PROSPERO and supported by the Wellcome Trust, UK National Institute for Health and Care Research, UK Foreign, Commonwealth and Development Office, and the NIH.

These findings have several important implications. First, in Southern and Eastern Africa the burden of untreated, bacteriologically-confirmed TB is markedly higher among people living with HIV, which underlines the continuing intersection of the two epidemics in those regions. Second, the higher N:P ratios for people living with HIV suggest that people with HIV who develop TB are more likely to be diagnosed and notified than people without HIV, implying relatively faster case detection among people living with HIV. Nonetheless, the N:P ratios for both groups and the overall findings show substantial delays and shortfalls in detecting untreated TB cases across the board. The researchers also point out a major gap in global evidence: nearly all eligible surveys came from Southern and Eastern Africa, leaving limited, community-representative data for other world regions where TB and HIV also cause harm. Overall, the study highlights both the elevated TB burden carried by people living with HIV in these African settings and the wider need to improve TB case finding and diagnosis for everyone, regardless of HIV status.