Out-of-hospital Respiratory Measures to Identify Patients With Serious Injury: A Systematic Review

Abstract

Objectives: The objective was to systematically review the published literature on the diagnostic accuracy of out-of-hospital respiratory measures for identifying patients with serious injury, focusing on measures feasible for field triage by emergency medical services personnel. Methods: We searched Ovid MEDLINE, CINAHL, and the Cochrane databases from January 1, 1996, through August 31, 2017. We included studies on the diagnostic accuracy (sensitivity, specificity, and area under the receiver operating characteristic curve [AUROC]) for all respiratory measures used to identify patients with serious injury (resource use, serious anatomic injury, and mortality). We assessed studies for risk of bias and strength of evidence (SOE). We performed meta-analysis for measures with sufficient data. Results: We identified 46 articles reporting results of 44 studies. Out-of-hospital respiratory measures included respiratory rate, pulse oximetry, and airway support. Meta-analysis was only possible for respiratory rate, which demonstrated a pooled sensitivity for serious injury of 13% (95% confidence interval [CI] = 5 to 29, I2 = 97.8%), specificity of 96% (95% CI = 83 to 99, I2 = 99.6%), and AUROC of 0.70 (95% CI = 0.66 to 0.79, I2 = 16.6%). For oxygen saturation, sensitivity ranged from 13% to 63%; specificity, 85% to 99%; and AUROC, 0.53 to 0.76. Need for airway support had a sensitivity of 8% to 53% and specificity of 61% to 100%; studies did not report AUROC. Across respiratory measures, the SOE was low. Other respiratory measures (pH, end-tidal carbon dioxide [CO2], and sublingual partial pressure of CO2) were reported only in emergency department studies. Conclusions: Data on the accuracy of out-of-hospital respiratory measures for field triage are limited and of low quality. Based on available research, respiratory rate, oxygen saturation, and need for airway intervention all have low sensitivity, high specificity, and poor to fair discrimination for identifying seriously injured patients.