Effect of different levels of PEEP on mortality in ICU patients without acute respiratory distress syndrome: systematic review and meta-analysis with trial sequential analysis

OBJECTIVE: To determine whether higher positive end- expiratory pressure (PEEP) could provide a survival advantage for patients without acute respiratory distress syndrome (ARDS) compared with lower PEEP. METHODS: Eligible studies were identified through searches of Embase, Cochrane Library, Web of Science, Medline, and Wanfang database from inception up to 1 June 2021. Trial sequential analysis (TSA) was used in this meta-analysis. DATA SYNTHESIS: Twenty-seven randomized controlled trials (RCTs) were identified for further evaluation. Higher and lower PEEP arms included 1330 patients and 1650 patients, respectively. A mean level of 9.6±3.4 cmH(2)O was applied in the higher PEEP groups and 1.9±2.6 cmH(2)O was used in the lower PEEP groups. Higher PEEP, compared with lower PEEP, was not associated with reduction of all-cause mortality (RR 1.03; 95% CI 0.91–1.18; P =0.627), and 28-day mortality (RR 1.07 ; 95% CI 0.92–1.24; P =0.365). In terms of risk of ARDS (RR 0.43; 95% CI 0.24–0.78; P =0.005), duration of intensive care unit (MD -1.04; 95%CI-1.36 to −0.73; P < 0.00001), and oxygenation (MD 40.30; 95%CI 0.94 to 79.65; P = 0.045), higher PEEP was superior to lower PEEP. Besides, the pooled analysis showed no significant differences between groups both in the duration of mechanical ventilation (MD 0.00; 95%CI-0.13 to 0.13; P = 0.996) and hospital stay (MD -0.66; 95%CI-1.94 to 0.61; P = 0.309). More importantly, lower PEEP did not increase the risk of pneumonia, atelectasis, barotrauma, hypoxemia, or hypotension among patients compared with higher PEEP. The TSA analysis showed that the results of all-cause mortality and 28-day mortality might be false-negative results. CONCLUSIONS: Our results suggest that a lower PEEP ventilation strategy was non-inferior to a higher PEEP ventilation strategy in ICU patients without ARDS, with no increased risk of all-cause mortality and 28-day mortality. Further high-quality RCTs should be performed to confirm these findings.