Evidence Supporting Clinical Use of Proportional Assist Ventilation: A Systematic Review and Meta-Analysis of Clinical Trials.

Background: While proportional assist ventilation (PAV), generates pressure in proportion to effort without a preselected target, proportional assist ventilation plus (PAV+) measures compliance and resistance, calculates work of breathing, and adjusts support to a preset assistance level. Objective: To summarize randomized controlled trials (RCTs) comparing invasive or noninvasive PAV or PAV+ in critically ill patients. Data Sources: We searched multiple databases to April 2017 without language restrictions and conference proceedings from 5 meetings to identify randomized parallel-group and crossover RCTs that compared invasive or noninvasive PAV or PAV+ to another mode in critically ill adults or children and reported at least 1 clinically important outcome. Results: We identified 14 RCTs (11 parallel group and 3 crossover) assessing PAV (n = 7) and PAV+ (n = 7) involving 931 adult patients. We found no effect of noninvasive PAV (vs noninvasive pressure support [PS]) on intubation (risk ratio 0.92 [0.59 to 1.43], I-2 = 0%) or invasive PAV (vs invasive PS) on percentage rapid eye movement sleep (mean difference [MD] -2.93% [-14.20 to +/- 8.34], I-2 = 43%). Compared to invasive PS, invasive PAV+ showed a nonsignificant increase in weaning time (MD +0.54 [-0.67 to +1.75] hours, I-2 = 0%), but no effect on hospital mortality, reintubation, or tracheostomy. Conclusions: Current evidence does not support the use of invasive or noninvasive PAV or invasive PAV+ in critically ill adults. Amid low to moderate heterogeneity, we identified 3 promising areas for future research including assessing the role of noninvasive PAV as an initial support strategy in patients with acute respiratory failure, invasive PAV on sleep quality during invasive ventilation, and possibly invasive PAV+ for weaning.