Association of Low-Dose Whole-Body Computed Tomography with Missed Injury Diagnoses and Radiation Exposure in Patients with Blunt Multiple Trauma

Abstract

Importance: Initial whole-body computed tomography (WBCT) for screening patients with suspected blunt multiple trauma remains controversial and a source of excess radiation exposure. Objective: To determine whether low-dose WBCT scanning using an iterative reconstruction algorithm does not increase the rate of missed injury diagnoses at the point of care compared with standard-dose WBCT with the benefit of less radiation exposure. Design, Setting, and Participants: This quasi-experimental, prospective time-series cohort study recruited 1074 consecutive patients admitted for suspected blunt multiple trauma to an academic metropolitan trauma center in Germany from September 3, 2014, through July 26, 2015, for the standard-dose protocol, and from August 7, 2015, through August 20, 2016, for the low-dose protocol. Five hundred sixty-five patients with suspected blunt multiple trauma prospectively received standard-dose WBCT, followed by 509 patients who underwent low-dose WBCT. Confounding was controlled by segmented regression analysis and a secondary multivariate logistic regression model. Data were analyzed from January 16, 2017, through October 14, 2019. Interventions: Standard- or low-dose WBCT. Main Outcomes and Measures: The primary outcome was the incidence of missed injury diagnoses at the point of care, using a synopsis of clinical, surgical, and radiological findings as an independent reference test. The secondary outcome was radiation exposure with either imaging strategy. Results: Of 1074 eligible patients, 971 (mean [SD] age, 52.7 [19.5] years; 649 men [66.8%]) completed the study. A total of 114 patients (11.7%) had multiple trauma, as defined by an Injury Severity Score of 16 or greater. The proportion of patients with any missed injury diagnosis at the point of care was 109 of 468 (23.3%) in the standard-dose and 107 of 503 (21.3%) in the low-dose WBCT groups (risk difference, -2.0% [95% CI, -7.3% to 3.2%]; unadjusted odds ratio, 0.89 [95% CI, 0.66-1.20]; P =.45). Adjustments for autocorrelation and multiple confounding variables did not alter the results. Radiation exposure, measured by the volume computed tomography dose index, was lowered from a median of 11.7 (interquartile range, 11.7-17.6) mGy in the standard-dose WBCT group to 5.9 (interquartile range, 5.9-8.8) mGy in the low-dose WBCT group (P