Systematic review of oxygenation and clinical outcomes to inform oxygen targets in critically ill trauma patients.

Abstract

Oxygen therapy is frequently administered to critically ill trauma patients to avoid hypoxia, but optimal oxygenation strategies are not clear. We conducted a systematic review of oxygen targets and clinical outcomes in trauma and critically ill patients. We searched Ovid MEDLINE, Cochrane Library, Embase, and Web of Science Core Collection from 1946 through 2017. Our initial search yielded 14,774 articles with 209 remaining after abstract review. We reviewed full text articles of human subjects with conditions of interest, an oxygen exposure or measurement, and clinical outcomes, narrowing the review to 43 articles. We assessed article quality using Grading of Recommendation, Assessment, Development, and Evaluation (GRADE) criteria. Of the 43 final studies meeting inclusions criteria, 17 focused on trauma and 26 studies focused on medical and/or surgical critical illness without trauma specifically. Four trauma studies supported lower oxygenation/normoxia, two supported higher oxygenation, and 11 supported neither normoxia nor higher oxygenation (five neutral and six supported avoidance of hypoxia). Fifteen critical illness studies supported lower oxygenation/normoxia, one supported higher oxygenation, and 10 supported neither normoxia nor higher oxygenation (nine neutral and one supported avoidance of hypoxia). We identified seven randomized controlled trials (four high quality, three moderate quality). Of the high-quality randomized controlled trials (none trauma-related), one supported lower oxygenation/normoxia and three were neutral. Of the moderate-quality randomized controlled trials (one trauma-related), one supported higher oxygenation, one was neutral, and one supported avoidance of hypoxia. We identified few trauma-specific studies beyond traumatic brain injury; none were high quality. Extrapolating primarily from nontrauma critical illness, reduced oxygen administration targeting normoxia in critically ill trauma patients may result in better or equivalent clinical outcomes. Additional trauma-specific trials are needed to determine the optimal oxygen strategy in critically injured patients. Systematic review, level IV.