Abstract
Background: Treatment recommendations are mostly built on RCTs with a methodology that ensure internal validity. However, their relevance may be compromised by the lack of representativity of study populations towards real-life patients.
Aims: We sought to identify and characterize the fraction of a real-life COPD population that would be eligible for inclusion into RCTs.
Methods: We used the Initiatives-BPCO database, a French cohort of 1309 well-characterized real-life COPD patients. A clinicaltrials.gov search identified phase 3 and 4 RCTs that included more than 500 COPD patients and used exacerbations related endpoints to assess medications. Eligibility criteria were extracted from each RCT and applied to the patients.
Results: The eligibility criteria of 18 RCTs were applied to the 1309 patients. The most discriminating eligibility criteria were FEV1, minimum exacerbation rate in the previous year and allergic status restrictions responsible for the exclusion of 52.1%, 41,6% and 28% of patients, respectively. Altogether, 2.3 to 46.7% of our patients would have satisfied all eligibility criteria, depending on the study.
Conclusions: These analyses confirm that an important gap exists between real-life patients and clinical trials populations in COPD, which limits the external validity of RCTs and therefore should be considered when grading levels of evidence and designing future studies.
- Copyright ©the authors 2017