Cardiovascular Risk Estimation Is Suboptimal in People With HIV.

Area of Science:

• Cardiology
• Infectious Diseases
• Epidemiology

Background:

• Established cardiovascular disease (CVD) risk prediction functions may not accurately assess risk in people with HIV.
• This study evaluated the performance of three common CVD risk prediction functions in two distinct HIV cohorts.

Purpose of the Study:

• To assess the accuracy of existing CVD risk prediction functions in people with HIV.
• To identify potential sex-based differences in the performance of these functions.

Main Methods:

• Utilized data from the Mass General Brigham and Kaiser Permanente Northern California HIV cohorts.
• Calculated CVD risk scores using the American College of Cardiology/American Heart Association (ACC/AHA) atherosclerotic CVD function, Framingham Heart Study (FHS) hard coronary heart disease function, and FHS hard CVD function.
• Assessed model performance through discrimination (c-statistics) and calibration, comparing predicted to observed risks stratified by sex.

Main Results:

• Across 9412 participants, 1.7% experienced a coronary heart disease event and 3.3% experienced a CVD event.
• All three functions generally underestimated CVD risk in women.
• In men, two functions underestimated risk, while one overestimated it. Calibration was poor for women with one function and for men with all functions.
• Discrimination was good for women (c-statistics 0.78-0.90) and moderate for men (c-statistics 0.71-0.72).

Conclusions:

• Existing CVD risk prediction functions tend to underestimate risk in people with HIV.
• Significant differences in model performance by sex highlight the necessity for both HIV-specific and sex-specific risk prediction tools.
• Developing tailored CVD risk prediction models for people with HIV is crucial for enhancing the care of aging individuals living with HIV.