The meaning and relevance of intervention trials in prevention

Abstract

A disease can be prevented theoretically in three different ways: by eradication, by avoiding the risk of acquiring and finally by managing, manipulating already established risk factors. In the case of atherosclerotic diseases only the third, least satisfactory approach is amenable to test the hypothesis of preventability by means of intervention trials. Prevention through correction of already established high risk is a fringe benefit of the purely medical approach to test the risk factor theory. The original postulate was simply: low risk profile goes with low incidence of the disease. Intervention trials contributed substantially to the broader acceptance of the preventability of atherosclerosis. The internal validity of the trial results is only one aspect of this problem, although this is the most often debated. The inference from or generalisability of positive and of course also of the negative trials poses problems only recently delt with in the literature. Issues of the sampling frame, exclusions, subjects eligible but not randomized, subgroup results, etc. and their influence on inference will be discussed, with the aim to redress the balance between randomised trials and good observational studies in prevention.