Multi-center Retrospective Study of Factors Affecting Perioperative Transfusion of Packed Red Blood Cells for Pelvic Fracture Patients.

objectiveTo analyze the use of packed red blood cells (PRBCs) for patients with pelvic fracture and evaluate factors associated with PRBC transfusion for patients with pelvic fracture.

methodsThis retrospective cohort study collected 551 patients with pelvic fractures from six hospitals between September 1, 2012, and June 31, 2019. The age span of patients varied from 10 to 95 years old, and they were classified into two groups based on high-energy pelvic fractures (HE-PFs) or low-energy pelvic fractures (LE-PFs). The study's outcome was the use of PRBCs, fresh frozen plasma (FFP), and albumin. Demographic data, characteristics, laboratory tests, clinical treatment details, and clinical outcomes were compared between the two groups. Factors that were statistically associated with perioperative PRBCs in univariate analyses were included to conduct an optimal scale regression to determine the independent factors for perioperative PRBCs.

resultsA total of 551 patients were screened from six hospitals, and after inclusion and exclusion, 319 were finally included and finished the follow-up from admission to discharge, while four patients died during hospitalization. Three hundred and nineteen patients were classified into two groups by their injury mechanisms. A total of 230/319 (72.1%) patients were classified into the HE-PF group, and 89/319 (27.8%) patients were classified into the LE-PF group. Patients in the HE-PF group were transfused with 4.5 (3-8) units of PRBCs, 300 (0-600) ml of FFP, and 0 (0-30) g of albumin, while patients in the LE-PF group were transfused with 3.5 (2-4.5) units of PRBCs, 0 (0-295) ml of FFP, and 0 (0-0) g of albumin (all P < 0.001). There were higher proportions of male patients and patients under 65 in the HE-PF group (all P < 0.001). HE-PF group patients were more severely injured and likely to take external fixation. The optimal scale regression revealed four significant factors associated with perioperative transfused PRBCs, which were patients on admission with hemorrhagic shock (importance = 0.283, P = 0.004), followed by fracture types identified by Tile classification (importance = 0.156, P < 0.001), hemoglobin levels below 70 g/L on admission (importance = 0.283, P = 0.004), followed by fracture types identified by Tile classification (importance = 0.156, P < 0.001), hemoglobin levels below 70 g/L on admission (importance = 0.148, P = 0.039), and methods of pelvic fixation (importance = 0.008, P = 0.026), ranked by the importance.

conclusionPatients with HE-PFs had increased transfusions of PRBCs, FFP, and albumin, and hemorrhagic shock on admission, Tile classification, Hb levels, and stabilization methods were found to be associated with perioperative PRBCs.