Skip Navigation
Department of Health and Human Services www.hhs.gov
 
Slide Tray
0 slides

Return to Slide Library

Slides

Add Presentation to Slide Tray Presentation:

Noninvasive Positive-Pressure Ventilation for Acute Respiratory Failure: Comparative Effectiveness

Slide: 21 of 25

Other Key Findings of the Comparative Effectiveness Review (3 of 3)

Effects of noninvasive positive-pressure ventilation (NPPV) on mortality and intubation rates appeared stronger in controlled clinical trials versus studies conducted in clinical practice settings. The strength of evidence for this finding was rated low. Of the 43 studies that compared NPPV to usual supportive care or invasive ventilation and reported mortality rates, 11 were efficacy trials, 2 were effectiveness trials, and 30 were mixed efficacy-effectiveness trials. The pooled odds ratios (ORs) for mortality were 0.56 (95-percent confidence interval [95% CI], 0.31–1.02) and 0.99 (95% CI, 0.66–1.49) for efficacy and effectiveness trials, respectively; for mixed efficacy-effectiveness trials, the pooled OR was 0.52 (95% CI, 0.41–0.66). An analysis that examined risk of intubation yielded similar results: for intubation, the pooled ORs for efficacy (n = 10), effectiveness (n = 2), and mixed efficacy-effectiveness trials (n = 25) were, respectively, 0.29 (95% CI, 0.19 to 0.46), 0.58 (95% CI, 0.16 to 2.13), and 0.29 (95% CI, 0.21 to 0.41). The difference in pooled ORs across the three categories of effectiveness was statistically significant for mortality (p = 0.02) but not for intubation (p = 0.61). It is important to note that for both mortality and intubation, there were only two effectiveness trials, which potentially render the pooled OR an unreliable estimate. Current evidence is therefore insufficient to estimate the effect of clinician experience, setting, system resources, and patient characteristics on treatment effects.

Limited evidence suggested that the treatment effects for NPPV on mortality and intubation rates were consistent across studies conducted in various geographical regions. The strength of evidence for this finding was rated low.

  • Meta-analyses of 39 studies that reported on mortality showed that the pooled OR for mortality for the 25 studies conducted in Europe was 0.58 (95% CI, 0.46 to 0.73), compared with 0.58 (95% CI, 0.25 to 1.33) for the 5 studies conducted in the United States or Canada, and 0.64 (95% CI, 0.36 to 1.13) for the 13 studies conducted in other countries. The difference in pooled ORs for mortality across the three categories of countries was not statistically significant (p = 0.95).

  • For the 39 studies that reported intubation rates, the pooled OR in meta-analyses was 0.33 (95% CI, 0.22 to 0.48) for the 19 studies conducted in Europe, 0.36 (95% CI, 0.20 to 0.66) for the 6 studies conducted in the United States or Canada, and 0.25 (95% CI, 0.14 to 0.43) for the 14 studies conducted in other countries. The difference in pooled odds ratios for intubation across the three categories of countries was not statistically significant (p = 0.70).