Predictive efficacy of Framingham's risk score in Indian scenario--a retrospective case-control study.

ABSTRACT

To assess the predictive ability of Framingham's risk score in primary prevention in our population, 252 cases and 212 age and sex matched controls were taken up for study. Those patients, who were presenting for the first time with acute coronarysyndrome (ACS) and who did not have any prior manifestations of coronary artery disease (CAD) and whose medical records were available formed the patient group. Framingham's risk score was calculated and the corresponding 10 years risk was assessed in each of them. The patients and controls were divided into two groups--diabetic and non-diabetic. Depending on the 10 years risk, they were further grouped into high risk (10 years risk > 20%), moderately high risk (10 years risk 10 to 20%) and low risk (10 years risk less than 10%). Results were compared and statistically analysed. In the diabetic patients with ACS 14% would have qualified as high risk, 33% as moderately high risk and 53% as low risk whereas in diabetic patients without any manifestation of CAD the distribution was 4% in the high risk, 54% in the moderately high risk and 42% in the low risk. In the non-diabetic subjects, amongst the patients of ACS, 20% would have been in high risk, 39% in moderately high risk and 41% in the low risk. The corresponding figures in the non-diabetic control subjects were 10% in high risk, 22% in the moderately high risk and 68% in the low risk. In the non-diabetic subjects, the mean risk was significantly more in patients than in controls (14.15% versus 8.61%, p <0.01). However, in the diabetic patients there was no significant difference in the mean projected risk between patients with ACS and patients without any manifestation of CAD (11.37% versus 10.41%, p>0.05).