The association between level of trauma care and clinical outcome measures: A systematic review and meta-analysis

BACKGROUND: With implementation of trauma systems, a level of trauma care classification was introduced. Use of such a system has been linked to significant improvements in survival and other outcomes. OBJECTIVES: The aim of this study was assessing the association between level of trauma care and fatal and nonfatal outcome measures for general and major trauma (MT) populations. METHODS: A systematic literature search was conducted using six electronic databases up to December 18, 2019. Studies comparing mortality or nonfatal outcomes between different levels of trauma care in general and MT populations (preferably Injury Severity Score of >15) were included. Two independent reviewers performed selection of relevant studies, data extraction, and a quality assessment of included articles. With a random-effects meta-analysis, adjusted and unadjusted pooled effect sizes were calculated for level I versus non–level I trauma centers. RESULTS: Twenty-two studies were included. Quality of the included studies was good; however, adjustment for comorbidity (32%) and interhospital transfer (38%) was performed less frequently. Nine (60%) of the 15 studies analyzing in-hospital mortality in general trauma populations reported a survival benefit for level I trauma centers. Level I trauma centers were not associated with higher mortality than non–level I trauma centers (adjusted odd ratio, 0.97; 95% confidence interval, 0.61–1.52). Of the 11 studies reporting in-hospital mortality in MT populations, 10 (91%) reported a survival benefit for level I trauma centers. Level I trauma centers were associated with lower mortality than non–level I trauma centers (adjusted odd ratio, 0.77; 95% confidence interval, 0.69–0.87). CONCLUSION: The association between level of trauma care and in-hospital mortality is evident for MT populations; however, this does not hold for general trauma populations. Level I trauma centers produce improved survival in MT populations. This association could not be proven for nonfatal outcomes in general and MT populations because of inconsistencies in the body of evidence.

• View at Publisher