Scoring system to predict hemorrhage in pelvic ring fracture

Abstract

Background: Risk factors for hemorrhage in patients with pelvic ring fracture have been widely reported. Because there are many risk factors, it is thought that prediction accuracy of hemorrhage in cases of pelvic ring fracture could be improved by using a scoring system.

Hypothesis: We investigated the risk factors for massive hemorrhage (MH) and created a novel predictive score of MH in pelvic ring fractures.

Material and methods: We retrospectively reviewed patients with pelvic ring fractures (Abbreviated Injury Score≥3 and age≥16 years) from January 2007 to June 2015. We excluded the cases that might have hemorrhage from other sites sufficient to require a blood transfusion. Massive hemorrhage was defined as hemorrhage requiring transfusion of≥6 red cell concentrate units within 24h of admission.

Results: The MH group included 27 patients and the non-MH group included 71 patients. Lactate level, AO/OTA classification and extravasation of computed tomography (CT) contrast fluid had a significantly higher risk as a result of multivariable analysis. The combined score using these risk factors according to their odds-adjusted ratios was created to predict for MH: lactate level>2.5-5.0 (mmol/L)=1 point,>5.0 (mmol/L)=2 points, partially stable (OA/OTA classification B1/B2/B3)=1 point, unstable (C1/C2/C3)=2 points, pelvic extravasation of contrast on CT=4 points. The AUC of the calculated score was 0.93 (95% CI: 0.89-0.98).

Conclusion: The combined score using these risk factors according to their odds-adjusted ratios was created to predict MH and was an effective prediction score.

Level of evidence: IV, retrospective study.

Keywords: Fractures; Hemorrhage; Pelvis; Prediction.