Evaluation of treatment benefit: randomized controlled trials and population-based observational research.

摘要

To the Editor: An editorial by Goodwin et al prioritized randomized controlled trials (RCTs) over observational research for demonstrating benefits of treatment. We agree that RCTs remain the gold standard for establishing efficacy of new cancer therapies under controlled conditions. However, we suggest that the editorial undervalued the ability of population-based observational research to investigate effects of new therapies among patients treated in routine practice. We disagree with the editorial statement that RCTs establish both efficacy and effectiveness and propose that population-based studies are necessary and complementary to ensure that results of RCTs translate into benefits for the general population, that is, to demonstrate effectiveness. The strength of RCTs derives from the power of randomization to ensure that the only difference between treatment arms is exposure to the treatment of interest. However, problems related to design, analysis, and reporting limit many RCTs and how they can be applied to manage patients in routine practice. Older patients, patients with comorbidities, and patients from disadvantaged socioeconomic backgrounds are underrepresented in RCTs, and there can be substantial differences in the provision of care for patients in RCTs compared with routine practice. Surrogate end points are used widely in RCTs despite lack of validation that they predict improvement in duration or quality of survival. RCTs underestimate and under-report harms from new cancer therapies, especially chronic toxicities in patients receiving targeted agents.