A nomogram-based HFpEF risk prediction score using age, sex, HIV status, smoking, hypertension, BMI, and BNP showed good discrimination (overfitting-corrected AUC 0.77) and calibration, offering a practical screening tool for primary care settings in sub-Saharan Africa with high HIV prevalence.
The overall prevalence of HFpEF in the study population was 7.8%, with a higher absolute proportion among people with HIV than those without HIV.
The study population was predominantly female, middle-aged, and had high rates of obesity and hypertension.
Two logistic regression models outperformed a random forest classifier for HFpEF risk prediction.
The final nomogram-based model demonstrated good discrimination and calibration for HFpEF risk prediction.
HIV status was included as an independent variable in the final risk prediction model, acknowledging HIV-related HFpEF risk not captured by existing tools.
The nomogram-based risk score was designed to support screening and targeted referral for HFpEF diagnosis in sub-Saharan African primary care settings.
This research suggests that a simple scoring tool can help identify people at high risk for a type of heart failure called heart failure with preserved ejection fraction (HFpEF) in primary care clinics in sub-Saharan Africa, where HIV is common. HFpEF is difficult to diagnose because it normally requires specialist cardiac imaging, which is not readily available in many primary care settings. The researchers developed their tool using data from over 1,500 adults in South Africa, including both people living with HIV on treatment and people without HIV, and found that about 8% of participants had HFpEF — with a slightly higher rate among people with HIV (8.4%) compared to those without (6.6%). The final risk score uses seven pieces of information that are easy to collect in a primary care visit: age, sex, HIV status, smoking history, whether the patient has hypertension, body mass index, and a blood test called BNP that can be done at the point of care. The tool performed well in statistical testing, with an accuracy score (AUC) of 0.77 after correction for overfitting, indicating good ability to distinguish between people who do and do not have HFpEF. Importantly, HIV status was deliberately included as a risk factor, addressing a gap in existing prediction tools that were not designed with HIV-positive populations in mind. This research suggests that such a nomogram-based score could be used by primary care providers to flag patients who should be referred for more specialized heart failure testing and care, without requiring specialist equipment or expertise at the first point of contact. This could be particularly valuable in high-HIV-prevalence settings in sub-Saharan Africa, where resources are limited but the burden of heart disease among people living with HIV is growing.
Honwana F, Wolfson J, Ahmed A, Huang Z, Mgidlana M, Aremu O, et al.. (2026). Development of a heart failure with preserved ejection fraction risk prediction score in primary care settings with a high HIV prevalence.. Open heart. https://doi.org/10.1136/openhrt-2026-004346