What are predictors of mortality in patients with pelvic fractures?

Abstract

Background: Our knowledge of factors influencing mortality of patients with pelvic ring injuries and the impact of associated injuries is currently based on limited information.

Questions/purposes: We identified the (1) causes and time of death, (2) demography, and (3) pattern and severity of injuries in patients with pelvic ring fractures who did not survive.

Methods: We prospectively collected data on 5340 patients listed in the German Pelvic Trauma Registry between April 30, 2004 and July 29, 2011; 3034 of 5340 (57%) patients were female. Demographic data and parameters indicating the type and severity of injury were recorded for patients who died in hospital (nonsurvivors) and compared with data of patients who survived (survivors). The median followup was 13 days (range, 0-1117 days).

Results: A total of 238 (4%) patients died a median of 2 days after trauma. The main cause of death was massive bleeding (34%), predominantly from the pelvic region (62% of all patients who died because of massive bleeding). Fifty-six percent of nonsurvivors and 43% of survivors were male. Nonsurvivors were characterized by a higher incidence of complex pelvic injuries (32% versus 8%), less isolated pelvic ring fractures (13% versus 49%), lower initial blood hemoglobin concentration (6.7 ± 2.9 versus 9.8 ± 3.0 g/dL) and systolic arterial blood pressure (77 ± 27 versus 106 ± 24 mmHg), and higher injury severity score (ISS) (35 ± 16 versus 15 ± 12).

Conclusion: Patients with pelvic fractures who did not survive were characterized by male gender, severe multiple trauma, and major hemorrhage.

Level of evidence: Level III, prognostic study. See Guidelines for Authors for a complete description of levels of evidence.

Fig. 1A–B

We used data from the German Pelvic Trauma Registry including 5340 patients from 31 medical centers (each institution is represented by a black column). (A) This graph shows the number of patients included by each center. (B) This graph shows the fraction of nonsurvivors in each center. There were no deaths with pelvic ring fractures in 9 institutions.

Fig. 2A–B

(A) This graph shows the causes of death in patients with pelvic fracture included injuries of the head (h), chest (c), abdomen (a), and pelvis (p). Most common general reasons not related to a specific body region were bleeding (b) and multiple organ failure (m) (highlighted in gray). Multiple reasons of death were possible. (B) This graph shows the origin of lethal bleeding included the head (h), chest (c), abdomen (a), and pelvis (p). Multiple origins of lethal bleeding were possible.

Fig. 3

The graph shows the survival curves of patients who died with pelvic ring fractures. More than 50% of patients who did not survive died within the first 2 days after trauma, while 3% of patients died later than 3 months after trauma.

Fig. 4

This graph shows the age distribution of inpatients who survived (survivors) versus inpatients who died (nonsurvivors) after pelvic ring fracture. The age distribution of nonsurvivors was comparable to that of survivors

Fig. 5A–B

(A) This graph shows Tile’s classification adopted by the Orthopaedic Trauma Association (OTA) [41] of pelvic ring fractures of patients who survived versus those who did not. Stable pelvic ring fractures were classified as Type A, fractures with only rotational instability as Type B, and fractures with both rotational and translational instability as Type C injuries. The fraction of Type C fractures was greater (*p < 0.001) in nonsurvivors than in survivors, while the fraction of Type A fractures was less (*p < 0.001) in nonsurvivors than in survivors. (B) This graph shows the mortality rate in patients with isolated pelvic fractures versus nonisolated pelvic fractures (multiple) stratified by the fracture type. The mortality rate was higher in patients with Type C fractures than in patients with Type B and A fractures, as well as in patients with nonisolated pelvic fractures than in patients with isolated pelvic fractures.

Fig. 6A–B

These graphs show the fraction of (A) complex pelvic injuries and (B) isolated pelvic fractures in patients who survived (survivors) versus patients who died (nonsurvivors) after pelvic ring fracture. The fraction of complex and nonisolated pelvic fractures (multiple) was greater (*p < 0.001) in nonsurvivors than in survivors, while the fraction of simple and isolated pelvic fractures was lesser (*p < 0.001) in nonsurvivors than in survivors.