961Absolute cardiovascular disease risk scores and medication use in rural India

Background Low-to-middle-income countries (LMICs) have limited resources to tackle the burden of cardiovascular disease (CVD). Most screening guidelines recommend the use of absolute risk scoring to determine treatment, but there is uncertainty among policy makers and clinicians about which risk algorithm to choose. We aimed to compare laboratory-based absolute CVD risk algorithms in a LMIC setting. Methods The study was conducted in the Rishi Valley, Andhra Pradesh, India. Over 8,000 participants were surveyed between 2012-2015. The 10-year absolute risk was computed and compared using the Framingham, WHO, and Australian absolute risk CVD algorithms. Results In participants aged 35-74 years, 151 (3%) had prior CVD. In all algorithms, absolute CVD risk increased with age and was greater in men than women. Using the WHO algorithm 4% were characterized as high-risk while >29% were at high-risk using the Australian risk tool. Agreement of risk classification among men ranged from a high of 84% (Spearman’s rho (rs) =0.92) between Australian and Framingham algorithms to 43% (rs=0.6) between the Australian and WHO risk scores. Among the high-risk population, only 15% were on lipid-lowering or antihypertensive therapy. Conclusions The Framingham and Australian risk scores enable some discrimination between high- and low-risk groups. However, the WHO algorithm underestimates these high-risk groups. Even though one third of the participants were at high-risk, most of them were not receiving recommended treatment. Key messages Lab-based CVD risk assessment tools have the potential in identifying high-risk populations in LMICs but the WHO risk scoring tool should be used with caution.