Rapid Retriage of Critically Injured Trauma Patients

Critically injured patients presenting to nontrauma hospitals require timely transfer to trauma centers^1,2; however, the transfer process varies and differences in outcomes for patients from trauma centers are unknown. We evaluated regional trauma outcomes after the local Emergency Medicine System implementation of policies that allow the expedited transfer of critically injured trauma patients or “rapid retriage” from nontrauma hospitals to trauma centers. After recognizing that a patient has critical traumatic injuries, emergency department (ED) physicians at a nontrauma hospital were given the ability to call 9-1-1 to retriage the patient to an associated trauma center with an unconditional acceptance, bypassing the standard transfer process. This study sought to determine differences in outcomes among (1) critically injured ED trauma transfer patients, (2) noncritically injured ED trauma transfer patients, and (3) critically injured trauma patients presenting directly to a trauma center within a regional trauma system. We hypothesized that improved outcomes for critically injured ED trauma transfer patients would be correlated with decreased transfer times and mechanism of penetrating injury.

This retrospective observational study used a quality improvement database compiled from all 12 trauma registries within the San Francisco Bay Area from January 1, 2013, to December 31, 2015. All trauma patients initially evaluated at a trauma center or transferred from a nontrauma hospital ED to a trauma center within the region were included (N = 49 438). Patients were categorized as (1) critically injured ED trauma transfer patients, (2) noncritically injured ED trauma transfer patients, and (3) critically injured trauma patients presenting directly to a trauma center. The triage criteria used to identify critically injured trauma patients were predefined as hypotension; a field Glasgow Coma Scale score less than 9 (score range: 3-15, with higher scores indicating better function); head injury with a dilated pupil; penetrating wound to the head, neck, back, thorax, or abdomen; open head wound; cardiac or respiratory distress; and the need for transfusion. The patient characteristics of Injury Severity Score (ISS) and mechanism of injury were reported. Primary outcomes were total transfer time and mortality. Secondary outcomes included time at a nontrauma hospital, interhospital transport time, admission rate, and trauma center length of stay. This study was deemed exempt by the Stanford Institutional Review Board because it was a quality improvement evaluation of data collected by trauma registrars for their institutional trauma registries. Thus, it did not meet the definition of human subject research.

Of the 49 438 patient records, 10 289 (20.8%) met the inclusion criteria (Table 1); of these 10 289 patients, 586 (5.7%) were critical ED trauma transfers. Compared with noncritically injured ED trauma transfer patients and critically injured trauma patients presenting directly to a trauma center, the critically injured ED trauma transfer patients had a higher median (interquartile range [IQR]) ISS of 13 (5-22) vs 9 (4-13) for the noncritically injured ED trauma transfer patients and 10 (3-25) for the critically injured trauma patients presenting directly to a trauma center. Critically injured ED trauma transfer patients also had a longer median (IQR) hospital length of stay: 5 (2-10) vs 3 (2-6) days for the noncritically injured ED trauma transfer patients and 4 (1-10) days for the critically injured trauma patients presenting directly to a trauma center. Mortality rate was similar among critically injured ED trauma transfer patients and those who presented directly to a trauma center: 16.6% (92 of 553 deaths) and 19.0% (1066 of 5598 deaths). However, critically injured ED trauma transfer patients who died spent more time at nontrauma hospitals than those who survived (143 vs 125 minutes).

Of 586 critically injured ED trauma transfer patients in whom the mechanism of injury was identified, 301 (51.4%) had blunt injuries and 285 (48.6%) had penetrating injuries. The patients with blunt injuries had a higher median (IQR) ISS (17 [10-26] vs 9 [2-16]; P < .001), higher mortality rate (23.5% vs 9.6%; P < .001), and longer median (IQR) time at a nontrauma hospital (160 [114-234] minutes vs 64 [36-125] minutes; P < .001) (Table 2).

Through this quality improvement process, we found that the rapid retriage protocol identified critically injured patients at higher risk for poor outcomes who could benefit from the use of rapid retriage. For critically injured ED trauma transfer patients who died, especially those with blunt trauma rather than penetrating trauma, increased time at a nontrauma hospital prior to transfer could have contributed to the higher mortality rate. In California’s exclusive trauma system, data for patients cared for in nontrauma hospitals are not systematically available, even for patients transferred to a trauma center. In this study, we identified important clinical data elements that are not routinely obtained from referring nontrauma hospitals. A regional trauma center committee is using this information to create a consensus-based trauma transfer performance improvement page within each site’s current trauma registry. This would enable the recording of data necessary to understand the emergency transfer process to improve the care for this vulnerable critically injured population.

Corresponding Author: N. Ewen Wang, MD, Department of Emergency Medicine, Stanford University School of Medicine, 900 Welch Rd, Ste 350, Third Floor, Palo Alto, CA 94304 (ewen@stanford.edu).

Accepted for Publication: April 16, 2017.

Published Online: July 5, 2017. doi:10.1001/jamasurg.2017.2178

Author Contributions: Ms Pirrotta and Dr Wang had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Della Valle, Kline, Spain, Pirrotta, Wang.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Della Valle, Pirrotta, Wang.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Kline, Pirrotta, Wang.

Administrative, technical, or material support: Della Valle, Kline, Pirrotta, Wang.

Study supervision: Newton, Spain, Wang.

Conflict of Interest Disclosures: None reported.