Methodological aspects in the design of secondary prevention trials

The aim of a secondary preventive trial is to produce results that may serve as a basis for therapeutic recommendations to other patients. The natural history of a disease studied including the mortality and reinfarction rate must be known and taken into consideration. The patients should be recruited without selection. By comparing the placebo mortality with expected levels the representativeness of patients can be assessed. One type of treatment can be expected to give different results in different groups of patients with the same disease, thus, prognostic prospective stratification may increase the value of comparisons and conclusions. The registration of end-points should preferably be done by a separate independent organization. Carefully classified specific mortality may be used as a major end-point in addition to total mortality. Similarly, different modes of deaths, e.g. sudden death, may be used if reliable definitions are used. Confounding factors are often difficult to isolate and identify and may have profound effects on the interpretation of a study. In all studies it is mandatory that the patient characteristics on entry do not differ between the different treatment groups. Concomitant treatment should be administered according to standardized criteria. The drop-out rate should be kept at a minimum. The possibility of generalization decreases with increasing drop-out rate. If the follow-up time becomes too long it is likely that at some time the relative benefit becomes less. Since the proportion of non-cardiovascular deaths increases with follow-up and age it may be critical to decide on the relevant follow-up time.