The generalisability of the results of RCTs investigating NIV in COPD

Introduction: The prevalence of chronic obstructive pulmonary disease (COPD) increases with age, and many patients have multiple comorbidities, particularly cardiovascular disease. Since the 1990s, non-invasive ventilation (NIV) has been studied as a therapeutic tool in COPD, however these trials often have narrow inclusion criteria. We sought to investigate whether patients in COPD NIV trials are representative of the general COPD population. Methods: All randomised controlled trials (RCTs) studying NIV in COPD were identified in a MEDLINE search by 2 reviewers. Non-randomised trials, and RCTs where the full text or English version were not available were excluded. Demographic data, inclusion and exclusion criteria were collated and compared with national COPD data and a large epidemiological study (Anecchino et al, International Journal of COPD 2007:2(4) 567-574). Results: Eighty trials were included (5338 patients, 67% male). Mean (SD) age was 67(4) years, and 18% of trials specifically excluded patients >75 or 80. There were no trials focussed on elderly patients. Ethnicity of patients was stated in only 1 paper. Fifty % of trials excluded patients with comorbidities; 45% excluded patients with cardiovascular disease. Comparison data shows COPD in the general population is most prevalent in the >70s, is 52% male and 68% have major comorbidities. Conclusion: Our review suggests that patients included in RCTs investigating NIV in COPD are not representative of our treatable patients. Older subjects and those with comorbidity are often excluded and women are less likely to be included. Future trials should seek to represent the COPD population more closely, to allow their results to be more generalisable.