Agreement between 10-year cardiovascular risk prediction models among people living with HIV in Kikwit, Democratic Republic of the Congo

People living with HIV have an increased burden of cardiovascular risk, but the cardiovascular risk prediction models used in clinical practice may generate substantially different estimates when applied to the same individuals. Evidence regarding the comparability of these models in sub-Saharan African populations living with HIV remains limited. This study compared the agreement between four cardiovascular risk prediction models and the World Health Organization (WHO) cardiovascular risk model among people living with HIV in Kikwit, Democratic Republic of the Congo. We conducted a retrospective cohort study among adults living with HIV receiving antiretroviral therapy at two HIV care centers in Kikwit between January 1 and June 30, 2026. Cardiovascular risk estimates were calculated using the Framingham, SCORE2, ASCVD, QRISK3, and WHO models. The WHO model was used as a comparative reference method because of its applicability to low- and middle-income settings, including sub-Saharan Africa, but was not considered a clinical gold standard. Agreement between each alternative model and the WHO model was assessed using Lin’s concordance correlation coefficient (CCC), Pearson’s correlation coefficient (ρ), bias-correction factor, mean differences with 95% limits of agreement using Bland–Altman analysis, the proportion of estimates within ± 30% of the WHO estimate (P30), and root mean squared error (RMSE). No prospective cardiovascular outcomes were available ; therefore, predictive performance, calibration, and discrimination against observed cardiovascular events were not assessed. A total of 107 people living with HIV were included. SCORE2 showed the highest concordance with the WHO model (CCC = 0.937; 95% CI: 0.909–0.957), with very high correlation (ρ = 0.997), a bias-correction factor of 0.940, and a mean difference of 1.72% points. Framingham also showed relatively high concordance (CCC = 0.818; 95% CI: 0.743–0.872), whereas ASCVD showed lower concordance (CCC = 0.807; 95% CI: 0.729–0.864) and a negative mean difference of − 2.86% points. QRISK3 showed the lowest concordance with the WHO model (CCC = 0.406; 95% CI: 0.235–0.554), lower correlation (ρ = 0.628), a mean difference of − 5.20% points, and substantially wider limits of agreement (− 21.57 to 11.17). The proportion of estimates within ± 30% of the WHO estimate was 92.5% for Framingham, 100.0% for SCORE2, 51.4% for ASCVD, and 9.3% for QRISK3. Among people living with HIV in this Kikwit cohort, SCORE2 showed the closest agreement with the WHO cardiovascular risk estimates, whereas QRISK3 showed the lowest concordance and greatest differences. These findings describe agreement between risk estimation models and should not be interpreted as evidence of superior prognostic accuracy. Prospective studies with longitudinal cardiovascular outcomes are needed to assess the calibration, discrimination, clinical utility, and comparative prognostic performance of cardiovascular risk prediction models in people living with HIV in the Democratic Republic of the Congo and other sub-Saharan African settings.