Virginia Triant, MD, MPH, Rajesh T. Gandhi, MD reviewing Feinstein MJ et al. JAMA Cardiol 2016 Dec 21.

Existing risk prediction scores were moderately accurate but underestimated the likelihood of events among important groups, such as blacks and individuals with low-to-moderate risk.

Accurate cardiovascular disease (CVD) risk prediction is essential to the care of HIV-infected individuals, who have an elevated risk for CVD with contributions from nontraditional HIV-related risk factors. In a U.S. cohort study, investigators assessed how well existing CVD risk prediction tools — the 2013 American College of Cardiology/American Heart Association Pooled Cohorts Equations (PCEs) for prediction of atherosclerotic cardiovascular disease (ASCVD) risk — function in HIV-infected adults, and whether including HIV-specific factors improves risk prediction.

In sex- and race-stratified analyses of 11,288 patients (82% men; >70% on antiretroviral therapy [ART]), the PCEs demonstrated adequate discrimination of MI risk overall but weaker discrimination among white women. Calibration, or model fit, was moderately strong overall but was weaker in groups other than white men. Observed MI event rates exceeded predicted rates in those with low-to-moderate risk (predicted ASCVD risk <10%) and particularly among black men and women. Risk scores that added HIV-specific factors (e.g., HIV RNA level, CD4 cell count, ART) did not improve on the PCEs and were less well calibrated.

CITATION(S):

Feinstein MJ et al. Assessing and refining myocardial infarction risk estimation among patients with human immunodeficiency virus: A study by the Centers for AIDS Research Network of Integrated Clinical Systems. JAMA Cardiol 2016 Dec 21; [e-pub]. 

JWatch