Arterial Embolisation for Trauma Patients with Pelvic Fractures in Emergency Settings: A Nationwide Matched Cohort Study in Japan

Objective: The objective of this study was to determine the association between arterial embolisation (AE) for pelvic fractures and death. Methods: The study had a retrospective design, using data from a nationwide population based prospective registry of trauma patients in Japan. This propensity score matched study included all adult patients from the registry with pelvic fractures between January 2004 and December 2018. The primary outcome was hospital death. Secondary outcomes included 28 day survival and length of hospital stay (LOS) in days. Multivariable logistic regression analyses were performed to control confounding variables, including patient, clinical, and hospital related variables; concomitant trauma; severe trauma; and haemodynamic instability. A conditional logistic regression analysis was performed to assess the association between treatment of pelvic fracture with AE and hospital mortality rate. Results: Among 17 670 eligible patients with pelvic fractures, 2 379 (13.5%) underwent AE (AE group) and 1 512 (8.6%) died in the hospital. After one to one propensity matching with 2 138 patients from each group (AE and non-AE), the hospital mortality rate was significantly lower in the AE group than in the non-AE group (15.0% vs. 18.1%; p = .007). The AE group had significantly lower mortality (odds ratio; 95% confidence interval [CI] 0.60; 0.43 - 0.84; p = .003) and a significantly higher 28 day mean survival rate than the non-AE group (0.89; 95% CI 0.87 - 0.90 vs. 0.86; 0.85 - 0.88; p = .003), although there was no significant difference in the LOS (48 days vs. 46 days; p = .11). Conclusion: This propensity score matched analysis showed an association between AE for pelvic fractures and lower hospital mortality rates. The findings in this large nationwide cohort study provide strong evidence for the benefit of embolisation for patients with pelvic fractures.