Objective The prehospital termination of resuscitation (TOR) guidelines for traumatic cardiopulmonary arrest (TCPA) was proposed in 2003. Its multiple descriptors of cases where efforts can be terminated make it complex to apply in the field. Here we proposed a simplified rule and evaluated its predictive performance.

Methods We analysed Utstein registry data for 2009–2013 from a Taipei emergency medical service to test a simplified TOR rule that comprises two criteria: blunt trauma injury and the presence of asystole. Enrolees were adults (≥18 years) with TCPA. The predicted outcome was in-hospital death. We compared the areas under the curve (AUC) of the simple rule with each of four descriptors in the guidelines and with a combination of all four to assess their discriminatory ability. Test characteristics were calculated to assess predictive performance.

Results A total of 893 TCPA cases were included. Blunt trauma occurred in 459 (51.4%) cases and asystole in 384 (43.0%). In-hospital mortality was 854 (95.6%) cases. The simplified TOR rule had greater discriminatory ability (AUC 0.683, 95% CI 0.618 to 0.747) compared with any single descriptor in the 2003 guidelines (range of AUC: 0.506–0.616) although the AUC was similar when all four were combined (AUC 0.695, 95% CI 0.615 to 0.775). The specificity of the simplified rule was 100% (95% CI 88.8% to 100%) and positive predictive value 100% (95% CI 96.8% to 100%). The false positive value, false negative value and decreased rate of unnecessary transport were 0% (95% CI 0% to 3.2%), 94.8% (95% CI 92.9% to 96.2%) and 16.4% (95% CI 14.1% to 19.1%), respectively.

Conclusions The simplified TOR rule appears to accurately predict non-survivors in adults with TCPA in the prehospital setting.