Comparative accuracy of different risk scores in assessing cardiovascular risk in Indians: a study in patients with first myocardial infarction.

BACKGROUND

Although a number of risk assessment models are available for estimating 10-year risk of cardiovascular (CV) events in patients requiring primary prevention of CV disease, the predictive accuracy of the contemporary risk models has not been adequately evaluated in Indians.

METHODS

149 patients [mean age 59.4 ± 10.6 years; 123 (82.6%) males] without prior CV disease and presenting with acute myocardial infarction (MI) were included. The four clinically most relevant risk assessment models [Framingham Risk score (RiskFRS), World Health Organization risk prediction charts (RiskWHO), American College of Cardiology/American Heart Association pooled cohort equations (RiskACC/AHA) and the 3rd Joint British Societies' risk calculator (RiskJBS)] were applied to estimate what would have been their predicted 10-year risk of CV events if they had presented just prior to suffering the acute MI.

RESULTS

RiskWHO provided the lowest risk estimates with 86.6% patients estimated to be having <20% 10-year risk. In comparison, RiskFRS and RiskACC/AHA returned higher risk estimates (61.7% and 69.8% with risk <20%, respectively; p values <0.001 for comparison with RiskWHO). However, the RiskJBS identified the highest proportion of the patients as being at high-risk (only 44.1% at <20% risk, p values 0 < 0.01 for comparison with all the other 3 risk scores).

CONCLUSIONS

This is the first study to show that in Indian patients presenting with acute MI, RiskJBS is likely to identify the largest proportion of the patients as at 'high-risk' as compared to RiskWHO, RiskFRS and RiskACC/AHA. However, large-scale prospective studies are needed to confirm these findings.