Positive end-expiratory pressure for preterm infants requiring conventional mechanical ventilation for respiratory distress syndrome or bronchopulmonary dysplasia
Journal Articles
Overview
Research
Identity
Additional Document Info
View All
Overview
abstract
BACKGROUND: Conventional mechanical ventilation (CMV) is a common therapy for neonatal respiratory failure. While CMV facilitates gas exchange, it may simultaneously injure the lungs. Positive end-expiratory pressure (PEEP) has received less attention than other ventilation parameters when considering this benefit-risk balance. While an appropriate PEEP level may result in clinical benefits, both inappropriately low or high levels may cause harm. An appropriate PEEP level may also be best achieved by an individualized approach. OBJECTIVES: 1. To compare the effects of PEEP levels in preterm infants requiring CMV for respiratory distress syndrome (RDS). We compare both: zero end-expiratory pressure (ZEEP) (0 cm H2O) versus any PEEP and low (< 5 cm H2O) vs high (≥ 5 cm H2O) PEEP.2. To compare the effects of PEEP levels in preterm infants requiring CMV for bronchopulmonary dysplasia (BPD). We compare both: ZEEP (0 cm H2O) vs any PEEP and low (< 5 cm H2O) versus high (≥ 5 cm H2O) PEEP.3. To compare the effects of different methods for individualizing PEEP to an optimal level in preterm newborn infants requiring CMV for RDS. SEARCH METHODS: We used the standard search strategy of Cochrane Neonatal to search the Cochrane Central Register of Controlled Trials, MEDLINE via PubMed, Embase, and CINAHL to 14 February 2018. We also searched clinical trials databases, conference proceedings, and the reference lists of retrieved articles for randomized controlled trials and quasi-randomized trials. SELECTION CRITERIA: We included all randomized or quasi-randomized controlled trials studying preterm infants born at less than 37 weeks' gestational age, requiring CMV and undergoing randomization to either different PEEP levels (RDS or BPD); or, two or more alternative methods for individualizing PEEP levels (RDS only). We included cross-over trials but limited outcomes to those from the first cross-over period. DATA COLLECTION AND ANALYSIS: We performed data collection and analysis according to the recommendations of the Cochrane Neonatal Review Group. We used the GRADE approach to assess the quality of evidence for prespecified key clinically relevant outcomes. MAIN RESULTS: Four trials met the inclusion criteria. Two cross-over trials with 28 participants compared different PEEP levels in infants with RDS. Meta-analysis was limited to short-term measures of pulmonary gas exchange and showed no differences between low and high PEEP.We identified no trials comparing PEEP levels in infants with BPD.Two trials enrolling 44 participants compared different methods for individualizing PEEP in infants with RDS. Both trials compared an oxygenation-guided lung-recruitment maneuver (LRM) with gradual PEEP level titrations for individualizing PEEP to routine care (control). Meta-analysis showed no difference between LRM and control on mortality by hospital discharge (risk ratio (RR) 1.00, 95% confidence interval (CI) 0.17 to 5.77); there was no statistically significant difference on BPD, with an effect estimate favoring LRM (RR 0.25, 95% CI 0.03 to 2.07); and a statistically significant difference favoring LRM for the outcome of duration of ventilatory support (mean difference -1.06 days, 95% CI -1.85 to -0.26; moderate heterogeneity, I2 = 67%). Short-term oxygenation measures also favored LRM. We graded the quality of the evidence as low for all key outcomes due to risk of bias and imprecision of the effect estimates. AUTHORS' CONCLUSIONS: There continues to be insufficient evidence to guide PEEP level selection for preterm infants on CMV for RDS or BPD. Low-quality data suggests that selecting PEEP levels through the application of an oxygenation-guided LRM may result in clinical benefit. Well-conducted randomized trials, particularly to further evaluate the potential benefits of oxygenation-guided LRMs, are needed.
