This study highlights two important points that distinguish it from other studies. First, to our knowledge, this report is the first to demonstrate that patients with trauma who experience TCA after physician interventions in the prehospital scenario are likely to have poor outcomes. Second, unlike previous studies, [1, 3, 4, 7,8,9,10, 14,15,16,17,18] this study used the registry data contributed by all HEMS institutions in Japan, thereby minimizing the deviations or biases in the patients’ characteristics and backgrounds.

In a recent review [1] of prehospital TCA cases, the mortality rate observed was 96.2% (95% CI: 95.0–97.2); when only the registry data were considered, the mortality rate was 97.2% (95% CI: 96.3–98.0). The rate of favorable neurological outcomes (CPC 1 or 2, or Glasgow Outcome Scale score 4 or 5) was 35.8%. Although it is challenging to provide direct comparisons with these results, our study showed a favorable mortality rate (93.8%) and favorable neurological outcomes (40.9%). Thus, the benefit of HEMS for TCA at the prehospital stage seemed consistent with the results of previous studies.

Prehospital TCA occurs in three phases. The first phase (the pre-EMS phase) occurs immediately after the patient is injured before the arrival of the EMS crew. In our study, the majority (60.3%) of prehospital TCA events occurred during this phase. However, the occurrence of TCA in the pre-EMS and EMS phases is not directly linked to poor outcomes. On the basis of our results, patients who experienced TCA in the pre-EMS and EMS phases had better prognoses than those who had TCA in the HEMS phase. During the prehospital TCA phase, patients who experienced TCA in the EMS phase had the most favorable outcome when the TCA was witnessed after the EMS team arrived and before the arrival of the HEMS team. This finding is consistent with that reported by Kitano et al. [19]. The most severe cases of prehospital TCA occurred in the HEMS phase, consistent with our clinical experience. However, one speculation is that the limited space and fewer personnel in the helicopter may hinder the provision of adequate chest compression and treatment for the patients during transport. From a different point of view, it might also be argued that the means to save the patient when cardiac arrest occurs in a situation where the highest prehospital intervention, the intervention of the physician, is limited.

In Japan, the EMS crew has a limited ability to provide care to patients with trauma before cardiac arrest; previous studies have shown that securing intravenous access did not improve the outcomes of patients with traumatic shock [12, 20]. However, Katayama et al. [20] reported that fluid administration by EMS reduced the incidence of cardiopulmonary arrest upon hospital arrival. The present study supports the hypothesis that intravenous access established by the EMS crew can increase the rate of ROSC. This might be related to the better rate of ROSC in the pre-EMS and EMS phases. Although epinephrine is a drug that the EMS crew can administer to patients with cardiac arrest, previous studies [14, 15, 19] and our study suggest that it may not improve the outcomes of patients with trauma.

Tranexamic acid, a drug that has not been previously studied in this context, may be effective against prehospital TCA in the future [21]. In a prehospital trauma care setting, not in patients with TCA, a recent study reported that tranexamic acid did not improve long-term neurological outcomes; however, it reduced 24-h mortality [22]. Although this study was not performed in patients with TCA, our study similarly showed the potential and efficacy of tranexamic acid in TCA. The EMS crews are not yet permitted to administer tranexamic acid in Japan; hence, further research is required.

Previous studies have reported conflicting evidence regarding the survival benefits of prehospital tracheal intubation for prehospital TCA [9, 16]. A recent review [1] found that tracheal intubation did not significantly affect the outcomes of patients with prehospital TCA, which was consistent with the findings of our study. However, the rate of sedative and muscle relaxant administration was higher during the HEMS phase, which may have contributed to the development of cardiac arrest. Although these drugs may be necessary for controlling agitation after achieving ROSC, they should be avoided during prehospital intubation of patients with severe trauma to prevent cardiac arrest. Moreover, bag-valve-mask ventilation is effective in prehospital situations; therefore, physicians should not necessarily prioritize the performance of tracheal intubation over this alternative [9].

Resuscitative thoracotomy is the most invasive procedure; although some studies have suggested that it may be effective for penetrating patients with trauma who are experiencing cardiac arrest [17, 23], others have reported poor outcomes associated with this procedure [18]. Our study supports the latter finding. In Japan, blunt trauma is more common than penetrating trauma, and there is limited evidence to support the use of resuscitative thoracotomy in patients with blunt trauma [24, 25]. Therefore, this procedure should not be performed in patients who experienced TCA without careful consideration of the underlying mechanism of injury.

Blood transfusion can improve the outcomes of patients with TCA [9]. However, it was not observed in the present study. This may be because prehospital blood transfusion is not yet widely performed by HEMS in Japan.

This study had some limitations. First, there was a considerable amount of missing data in the JSAS-R, especially for TCA situations, which made the analysis difficult. Second, data on the detailed timing of the cardiac arrest and ROSC, first monitored rhythm, and presence of bystander cardiopulmonary resuscitation were not registered in the JSAS-R database. Third, there was a potential for patient selection bias when using the registry data. Fourth, as this was a retrospective study, the causality between patient outcomes and medical interventions could not be proven. Fifth, this database did not register the timing of interventions, so we could not identify whether the medical interventions had been performed as resuscitation, performed after resuscitation, or performed for other purposes. Finally, the causes of TCA were not identified.