Key points are not available for this paper at this time.
Post-extubation respiratory failure requiring reintubation is associated with substantial morbidity and mortality in critically ill patients. The comparative effectiveness of available non-invasive respiratory support strategies, conventional oxygen therapy (COT), high-flow nasal cannula (HFNC), non-invasive ventilation (NIV), and the combination of NIV and HFNC (NIV+HFNC), remains debated. This systematic review and network meta-analysis (NMA) aimed to determine the comparative effectiveness of post-extubation respiratory support modalities in preventing reintubation in critically ill adults. MEDLINE, Embase, the Cochrane Central Register of Controlled Trials (CENTRAL), the Cumulative Index to Nursing and Allied Health Literature (CINAHL), and Scopus were searched from 1 January 2010 to 31 May 2026. Randomized controlled trials (RCTs) and adjusted non-randomized comparative studies evaluating post-extubation respiratory support applied within 24 hours of planned extubation in adult intensive care unit (ICU) patients were eligible for inclusion. Twenty-eight studies (22 RCTs and six observational studies) met the inclusion criteria. The NMA demonstrated that all active respiratory support strategies significantly reduced the risk of reintubation compared to COT. The combined NIV+HFNC strategy was ranked as the most effective intervention (P-score=0.81; risk ratio (RR) vs. COT: 0.53 (95% CI: 0.33-0.84); low certainty), followed closely by NIV alone (P-score=0.79; RR vs. COT: 0.55 (95% CI: 0.41-0.76); low certainty) and HFNC alone (P-score=0.40; RR vs. COT: 0.68 (95% CI: 0.52-0.89); moderate certainty). There were no statistically significant differences in reintubation rates between the active interventions (NIV+HFNC vs. HFNC: RR 0.78 (0.50-1.22); NIV vs. HFNC: RR 0.81 (0.63-1.05)). In critically ill adults undergoing planned extubation, active non-invasive respiratory support (HFNC, NIV, or their combination) reduces the risk of reintubation compared to COT. The sequential combination of NIV and HFNC ranked highest probabilistically, though direct comparisons between active interventions lacked statistical significance. The choice of support should be individualized according to patient risk profile, clinical context, tolerance, and certainty of available evidence.
Abdeltawab et al. (Wed,) studied this question.