Objective: To estimate the effect of non-invasive mechanical ventilation (NIV) on the rate of reintubation among patients undergoing cardiothoracic surgery.
Design: A meta-analysis of randomised trials.
Data sources: Medline, Embase, and the Cochrane Central Register of clinical trials were searched (April 2012) for pertinent studies by two trained investigators. International experts were contacted.
Data extraction: Articles were assessed by two trained investigators, with divergences resolved by consensus. Inclusion criterion was random allocation to NIV versus standard treatment without restrictions on duration or modalities of the treatment delivered.
Data synthesis: Fourteen studies enrolling 1211 patients were included in the meta-analysis. NIV reduced the reintubation rate (risk ratio [RR], 0.29; 95% CI, 0.16-0.53; P for efficacy < 0.0001; I2 = 0), hospital length of stay and mortality. Subgroup analyses suggested that the benefits of NIV are more important in patients with ongoing acute respiratory failure (RR, 0.25; 95% CI, 0.07-0.89) and in those at high risk of developing postoperative pulmonary complications (RR, 0.19; 95% CI, 0.04-0.84). Analyses including prophylactic studies in patients at low risk did not show a significant effect of NIV on reintubation rate (RR = 0.42; 95% CI, 0.12-1.48) and on any of the outcomes considered except for oxygenation.
Conclusions: NIV seems to be effective in reducing reintubation rate after cardiothoracic surgery. The results of this meta-analysis should be confirmed by large randomised controlled studies.