Is there a role for real life observational studies in integrating knowledge provided by randomized clinical trials?

Abstract

Randomized controlled trials (RCTs) are the gold standard for determining the efficacy of treatment agents, because they represent the only secure method of attributing causality in observed associations between treatments and outcomes. The main strengths of RCTs are well known to researchers. In particular: 1. they reduce the risk of selection bias. The random allocation of patients to treatment or control groups balances their measured and unmeasured characteristics; 2. they allow assessment of the cause-effect association. The impact of treatment is compared with the impact of placebo in the control group. But researchers are also aware of some important limitations of RCTs. In particular: 1. strict inclusion and exclusion criteria for patient selection are always applied, in order to eliminate factors which may confound or obscure treatment effects. However, patient selection limits the generalizability of results, and clinicians recognize that many of the patients they see in routine clinical practice, and for whom they must make therapeutic decisions, would not be eligible for inclusion in RCTs. Therefore, the main issue raised is the applicability of trial results to ordinary patients; 2. most RCTs involve younger adults, and treatment decisions in elderly patients are derived from them. The benefit-to-risk ratio of any given intervention may be quite different in frail older patients with significant comorbidities, and the applicability of study Is there a role for real life observational studies in integrating knowledge provided by randomized clinical trials?