This article is more than 5 years old. Information may no longer be current.
Most prediction equations inaccurately assess CVD risk in HIV
Click Here to Manage Email Alerts
Current models for predicting the risk for cardiovascular disease in patients with HIV were inaccurate, according to an analysis from the ongoing HIV Outpatient Study.
Researchers compared three common models used for predicting cardiovascular disease (CVD) in the general population: The Framingham general cardiovascular Risk Score (FRS), the American College of Cardiology/American Heart Association Pooled Cohort Equations (PCE) and the Systematic COronary Risk Evaluation (SCORE) high-risk equation — along with the Data Collection on Adverse Effects of Anti-HIV Drugs (D:A:D) study equation, which is intended to predict risk for CVD among patients with HIV.
“Only the FRS accurately estimated the risk of CVD events, while PCE and D:A:D underestimated risk,” Angela M. Thompson-Paul, PhD, MSPH, of the CDC’s Division of HIV/AIDS Prevention, Epidemiology and Surveillance Branch, and colleagues wrote. “Although these models could potentially be used to rank U.S. HIV-infected individuals at higher or lower risk for CVD, the models may fail to identify substantial numbers of HIV-infected persons with elevated CVD risk who could potentially benefit from additional medical treatment.”
Thompson-Paul and colleagues identified patients from the HIV Outpatient Study who had attended at least two appointments with researchers from January 2002 to Oct. 1, 2010 (n = 2,283). The researchers noted that 50% of these patients were classified as overweight or obese, 47.8% had hypertension and 16.9% had hypercholesterolemia. Another 13.3% of the cohort had been prescribed statins. About one-third had experienced a prior AIDS-defining illness, and 86.9% used combination ART. Patients had a median nadir CD4 count of 211 cells/mm3.
Over the course of 15,056 person-years, 8.5% of patients experienced a CVD event, Thompson-Paul and colleagues reported. The researchers wrote that the FRS was well-calibrated and showed moderate discrimination (C-statistic: 0.66; expected/observed events [E/O]: 1.01), while the PCE demonstrated good discrimination, but was not as well-calibrated (C-statistic: 0.71; E/O: 0.88; P < .001). The D:A:D performed similarly to the PCE (C-statistic 0.72; E/O: 0.8; P < .001). Thompson-Paul and colleagues reported that the SCORE performed the poorest of the models (C-statistic: 0.59; E/O: 1.72).
“In conclusion, current CVD risk prediction equations were able to adequately distinguish U.S. HIV-infected adults with higher or lower CVD risk, but inaccurately predicted CVD events. Among those at highest CVD risk, less than half reported statin and/or aspirin use,” Thompson-Paul and colleagues wrote. “These models may fail to identify substantial numbers of HIV-infected persons with elevated CVD risk who could potentially benefit from additional medical treatment.” – by Andy Polhamus
Disclosure: Thompson-Paul reports no relevant financial disclosures. Please see the full study for a list of all other authors’ relevant financial disclosures.
Click Here to Manage Email Alerts