Aim: To compare the Framingham and Prospective Cardiovascular Munster (PROCAM) risk calculations.
Methods: We calculated the risk in 234 dyslipidaemic patients without overt vascular disease and in different subgroups. For example, the proportion of patients with coronary heart disease (CHD) risk >or= 20%, the effect of including the family history (FaHist) and of adjusting raised triglyceride (TG) levels.
Results: The Framingham risk was significantly (p < 0.0001) higher than the PROCAM risk (with and without including the FaHist) in different subgroups and when the TGs were adjusted to 1.7 mmol/l. The percentage of patients with CHD risk >or= 20% calculated by the Framingham (based on systolic or diastolic blood pressure) and PROCAM equations was 21.4% or 23.1% and 16.2% respectively. In the tertile with the highest PROCAM risk, the Framingham score was significantly greater than the PROCAM risk only when the FaHist was included in the Framingham calculation. When we analysed risk by gender, the Framingham score did not differ but the PROCAM risk was significantly (p < 0.0001) greater in men. When TG values were adjusted to 1.7 mmol/l, the predicted risk using PROCAM changed by 0% to -2% in all subgroups.
Conclusions: In dyslipidaemic patients without overt vascular disease the Framingham model predicted a higher risk than PROCAM. Thus, the Framingham equation probably leads to substantial overtreatment compared with PROCAM. However, according to the literature, even the PROCAM equation may overestimate risk. This has considerable cost implications. New more accurate risk engines are needed to calculate risk in dyslipidaemic patients without overt vascular disease.