Background/introduction The Denver HIV risk score (DHRS) is a prediction rule developed for targeting HIV testing and validated in U.S. clinical settings (PMID: 22431561). The final model of the DHRS included age, gender, race/ethnicity, sex with a male, vaginal intercourse, receptive anal intercourse, injection drug use, and past HIV testing.

Aim(s)/objectives We aimed to validate the DHRS in patients attending two publicly funded STI clinics in Vancouver, British Columbia.

Methods We validated the model using electronic records (2000–2012) from 47,175 clinic visits. Each visit was scored based on variables included in the DHRS. Visits were stratified into 5 risk groups according to their score: very low (<20), low (20–29), moderate (30–39), high (40–49), and very high (≥50). The model’s discrimination and calibration for predicting an HIV diagnosis were examined by AUC and the Hosmer-Lemeshow (H-L) statistic. We examined the sensitivity and proportion of patients that would need to be screened at different cutoffs of the risk score.

Results The prevalence of HIV infection was 0.46%. Validation demonstrated good performance: the AUC was 0.80 (95% CI: 0.79–0.81) and the H-L χ² = 8.8, 8 df, p = 0.36. HIV prevalence within each risk groups was: 0%, 0.05%, 0.25%, 0.86%, and 1.23%, respectively. HIV testing is recommended for scores of ≥40. The DHRS identified cases with a sensitivity of 96% and a fraction screened of 41%.

Discussion/conclusion The DHRS performed well in these STI clinic settings in Vancouver, accurately identifying individuals at increased HIV risk, and may be useful for providing individualised estimates of risk as part of routine HIV screening.