Assessing HIV vulnerability among young women in South Africa: A comparison of three risk assessment tools

Background: Young women in sub-Saharan Africa remain disproportionately affected by HIV. However, self-perceived HIV risk may not align with reported sexual behaviours or other dimensions of risk. Behavioural measures, such as the Balloon Analogue Risk Task (BART), may capture risk-taking propensity not reflected in conventional self-report measures. We compared self-perceived HIV risk, behaviour-based HIV risk, and behavioural risk-taking propensity among young women in KwaZulu-Natal, South Africa. Methods: We conducted a cross-sectional survey between July and November 2025 among self-reported HIV-negative women aged 18-30 years in KwaZulu-Natal, South Africa. Self-perceived HIV risk was assessed using the Perceived Risk of HIV Scale (PRHS), behaviour-based HIV risk using the Vaginal and Oral Interventions to Control the Epidemic (VOICE) Risk Score, and behavioural risk-taking propensity using BART. Spearman's rank correlation coefficients were used to examine convergence between the three measures. Results: Among 250 participants, median age was 22 years (IQR 20-25), all identified as Black, and 53.6% were attending school or college. Overall, 42.0% were classified as high risk by VOICE, 6.0% by PRHS, and 33.1% of participants with complete BART data were in the highest tertile of risk-taking propensity. Mean age differed across VOICE categories (p = 0.036), but not PRHS (p = 0.111) or BART categories (p = 0.698). PRHS categories were associated with anal intercourse, oral intercourse, consistent condom use, ever having sold sex, and age-disparate partnerships (all p < 0.05). VOICE categories were not significantly associated with the assessed sexual behaviours, while BART categories were associated with vaginal intercourse in the past month (p = 0.042). Correlations between PRHS, VOICE, and BART were weak and non-significant (Spearman's rho = 0.001-0.057; all p > 0.05). Conclusions: Self-perceived HIV risk, behaviour-based HIV risk, and behavioural risk-taking propensity showed limited convergence in this sample. The measures also differed in their associations with participant characteristics and reported sexual behaviours, suggesting that they capture distinct dimensions of risk. Further research should investigate whether integrating behavioural measures with conventional self-report assessments improves understanding of HIV vulnerability and informs prevention approaches tailored to young women.