A network meta-analysis of different positive airway pressure interventions in obesity hypoventilation syndrome

Continuous positive airway pressure (CPAP) and different types of non-invasive ventilation (NIV) have been studied in obesity hypoventilation syndrome such as bi-level PAP with back-up rate (BPAP-BUR), BPAP without BUR, and the new hybrid devices that target a pre-set volume by adjustment of pressure support (VT-PS). Although several studies have compared one PAP intervention with the other, none has compared all four in a head-to-head design, which formed the basis of this network meta-analysis. PubMed and Web of Science were searched for potentially includable randomised active comparator trials. Changes in partial pressure of carbon dioxide (PaCO₂ ) and Epworth Sleepiness Score (ESS) were the primary outcomes of interest. Network meta-analysis was done in R program using the 'frequentist' framework. A total of seven trials were included. Only VT-PS and BPAP-BUR showed statistically significant reductions in PaCO₂ compared to control, with no significant inter-PAP differences except for the comparison between VT-PS and CPAP. Only VT-PS showed a statistically significant improvement in ESS as compared to control, with no other significant inter-PAP differences. P-score ranking (based on effect size and standard errors) and Hasse diagram ranked VT-PS and BPAP as superior to other PAPs for both primary outcomes. There were no significant differences between the different PAP interventions for hospital or emergency department admissions. The results of this network meta-analysis suggest superiority of VT-PS and BPAP over other PAP interventions at least for daytime hypercapnia and subjective daytime somnolence.