Predictors of Informed People’s Preferences for Statin Therapy to Reduce Cardiovascular Disease Risk: an Internet Survey Study

Abstract

Background: Guidelines suggest clinicians inform patients about their 10-year cardiovascular disease (CVD) risk; however, little is known about how the risk estimate influences patients’ preferences for statin therapy for primary prevention. Objective: To define predictors of preference for statin therapy after participants were informed about their individualized benefits and harms. Design: Cross-sectional survey in 2020. Setting: Online US survey panel. Participants: A national sample of 304 respondents aged 40 to 75 who had not previously taken a statin and who knew their cholesterol levels and blood pressure measurements. Intervention: Participants entered their risk factors into a calculator which estimated their 10-year CVD risk. They were then provided with an estimate of their absolute risk reduction with a statin and the chance of side effects from meta-analyses. Main Measurements: We used a hierarchical model to predict participants’ preferences for statin therapy according to their 10-year CVD risk, perceptions of the magnitude of statin benefit (large, medium, small, or almost no benefit), worry about side effects (very worried, somewhat worried, a little worried, not worried at all), and other variables. Key Results: Participants had a mean age of 55 years (SD = 9.9); 50% were female, 44% were non-white, and 16% had a high school degree or less education. After reviewing their benefits and side effects, 45% of the participants reported they probably or definitely wanted to take a statin. In the full hierarchical model, only perceived benefits of taking a statin was a significant independent predictor of wanting a statin (OR 7.3, 95% CI 4.7, 12.2). Limitations: Participants were from an internet survey panel and making hypothetical decisions. Conclusions: Participants’ perceptions of their benefit from statin therapy predicted wanting to take a statin for primary prevention; neither estimated CVD risk nor worries about statin side effects were independent predictors.