Inadequate Statistical Power of Published Comparative Cohort Studies on Ventilator‐Associated Pneumonia to Detect Mortality Differences

BACKGROUND: Comparative cohort studies are often conducted to identify novel therapeutic strategies or prognostic factors for ventilator-associated pneumonia (VAP). We aimed to evaluate the power of such studies to provide clinically and statistically significant conclusions with regard to mortality differences. METHODS: We searched in PubMed and Scopus for comparative cohort studies that evaluated mortality in patients with VAP. We calculated the central estimates and corresponding 95% confidence intervals (CIs) for mortality differences between compared patient groups. We also calculated the statistical power of the included studies to detect a difference in mortality that corresponds to a risk ratio of 0.80. RESULTS: We identified 39 (20 prospective) comparative cohort studies on VAP as eligible for inclusion in this analysis. The median absolute risk difference in mortality between compared groups was 10% (interquartile range IQR, 5%-18%), and the median width of the 95% CI of the absolute risk difference in mortality was 34% (IQR, 28%-42.5%). The median power of the included studies to detect a risk ratio for mortality of 0.80 was 14.7% (IQR, 10.6%-21.8%). CONCLUSIONS: There is considerable uncertainty around the central estimate of comparative cohort studies on VAP with regard to mortality differences. For a wiser use of resources allocated to research, we emphasize the need to conduct cohort studies with larger sample size so that potential differences between the compared groups are more likely to be shown.