Imposter or knight in shining armor? Pelvic circumferential compression devices (PCCD) for severe pelvic injuries in patients with multiple trauma: a trauma-registry analysis

Abstract

Abstract Background Pelvic Circumferential Compression Devices (PCCD) are standard in hemorrhage-control of unstable pelvic ring fractures (UPF). Controversial data on their usefulness exists. Aim of the study was to investigate whether prehospital application of PCCD can reduce mortality and transfusion requirements in UPF. Methods Retrospective cohort study. From 2016 until 2021, 63,371 adult severely injured patients were included into TraumaRegister DGU® of the German Trauma Society (TR-DGU). We analyzed PCCD use over time and compared patients with multiple trauma patients and UPF, who received prehospital PCCD to those who did not (noPCCD). Groups were adjusted for risk of prehospital PCCD application by propensity score matching. Primary endpoints were hospital mortality, standardized mortality rate (SMR) and transfusion requirements. Results Overall UPF incidence was 9% (N = 5880) and PCCD use increased over time (7.5% to 20.4%). Of all cases with UPF, 40.2% received PCCD and of all cases with PCCD application, 61% had no pelvic injury at all. PCCD patients were more severely injured and had higher rates of shock or transfusion. 24-h.-mortality and hospital mortality were higher with PCCD (10.9% vs. 9.3%; p = 0.033; 17.9% vs. 16.1%, p = 0.070). Hospital mortality with PCCD was 1% lower than predicted. SMR was in favor of PCCD but failed statistical significance (0.95 vs. 1.04, p = 0.101). 1,860 propensity score matched pairs were analyzed: NoPCCD-patients received more often catecholamines (19.6% vs. 18.5%, p = 0.043) but required less surgical pelvic stabilization in the emergency room (28.6% vs. 36.8%, p < 0.001). There was no difference in mortality or transfusion requirements. Conclusion We observed PCCD overuse in general and underuse in UPF. Prehospital PCCD appears to be more a marker of injury severity and less triggered by presence of UPF. We found no salutary effect on survival or transfusion requirements. Inappropriate indication and technical flaw may have biased our results. TR-DGU does not contain data on these aspects. Further studies are necessary. Modular add-on questioners to the registry could offer one possible solution to overcome this limitation. We are concerned that PCCD use may be unfairly discredited by misinterpretation of the available evidence and strongly vote for a prospective trial.