Evaluating clinical systems: the social programme perspective

Abstract

Randomized Controlled Trials (RCTs) are seen as the ‘gold standard’ for the evaluation of mature clinical systems. The author contests this viewpoint and argues for an alternative approach derived from the evaluation of social programmes. RCTs assume the experimental system is a black box, the emphasis being to establish the presence of a causal link between the intervention and outcome, but not to understand the mechanism by which the outcome occurs. Conversely, the social programme perspective recognizes explicitly that for changes in outcome to occur, a mechanism and sustaining context is needed, and that both of these can, and should be, studied in detail. This perspective causes a subtle shift in the emphasis and experimental approach of an evaluation, and can provide valuable information to assist a major goal in medical informatics; to engineer useful and useable systems and devise strategies to ensure they are successfully deployed into routine clinical practice.