In Reply The issue of combining randomized clinical trials and nonrandomized studies in a meta-analysis is a common challenge faced during evidence synthesis. When an abundance of well-done randomized clinical trials does not exist, clinical decision-making has to proceed based on the available evidence. In our systematic review1 about home noninvasive positive pressure ventilation (NIPPV) in COPD, we judged that combining these 2 types of studies was appropriate. For transparency purposes, we presented estimates separately from each study design. When examining the 2 designs, we observed similar trends, consistent results (overlapping CIs of relative estimates of mortality, need for intubation, and hospitalization), and no significant heterogeneity. A methodology study by the Cochrane Collaboration that examined 1583 meta-analyses covering 228 different medical conditions concluded that there was little evidence for significant effect estimate differences between observational studies and randomized clinical trials (ratio of odds ratios, 1.08; 95% CI, 0.96-1.22).2 This should not be interpreted as equating inferences from the 2 designs, and there are many examples of misleading observational studies. It simply means that the totality of evidence needs to be considered when not enough trials are available.
Wilson ME, Murad MH, Wang Z. Home Noninvasive Positive Pressure Ventilation for Chronic Obstructive Pulmonary Disease—Reply. JAMA. 2020;323(23):2432–2433. doi:10.1001/jama.2020.6093
Customize your JAMA Network experience by selecting one or more topics from the list below.
Create a personal account or sign in to: