Vist GE et al. (2005) Systematic review to determine whether participation in a trial influences outcome. BMJ 330: 1175

A systematic review published in the British Medical Journal has asked whether participation in randomized controlled trials (RCTs) influences clinical outcomes. By comparing outcomes of RCT participants with those of individuals receiving similar treatments outside of trials, the researchers demonstrated that RCTs were associated with neither a harmful nor a beneficial effect.

There has been some debate over whether participation in a randomized trial increases a patient's risk of a bad outcome, and whether the benefit of such studies is restricted to future patients. Furthermore, the applicability of clinical trial results to normal clinical practice has been questioned. Several studies have been undertaken to address these questions, but no single study has provided conclusive evidence. Vist et al. carried out a systematic review of 50 non-randomized cohort studies, and five clinical trials in which patients were randomized to participation or the option of participation. The studies were in the fields of oncology, obstetrics and gynecology, cardiology or other internal medicine, psychology and drug misuse, pediatrics, and respiration. The studies included data on a total of 31,140 RCT participants and 20,380 comparable non-participants who received similar treatment.

Of 73 dichotomous main outcomes studied, the majority (59) showed no statistically significant difference between RCT participants and non-participants, whereas 10 showed significantly better outcomes and 4 showed significantly worse outcomes for RCT participants. The pooled results for 18 continuous outcomes showed no statistically significant difference between those treated inside or outside of RCTs.

In summary, this systematic review indicates that participation in RCTs is associated with neither benefit or harm, compared with the use of similar interventions outside trials. These findings support the idea that the results of RCTs are generally applicable to clinical practice.