Methods of assessing medical treatments

BMJ 1994;309:741 (17 September)

Letters

Methods of assessing medical treatments

EDITOR, - We agree with Trevor A Sheldon that randomised clinicaltrials are one of the most important methods of assessing themerit of medical treatments.¹ We question, however, whetherthey can, by themselves, address successfully two critical challengesthat Sheldon identifies: "to provide answers to more clinicallyrelevant questions" and "to get the results of research intopractice."

Quality of care is the extent to which health care is able toachieve those health benefits that science and technology makepossible,² or the gap between efficacy - the probability ofbenefits under ideal conditions - and effectiveness.³ Whetherdrugs or techniques of proved efficacy are used effectivelydepends largely on factors related to the clinical setting andthe clinician's practice. Additionally, achieving desired healthoutcomes often requires the active participation of the patient.²Research focusing on efficacy seeks to exclude such factors,either by design or by statistical analysis. . . . [Full text of this article]