Factors of Pelvic Infection and Death in Patients with Open Pelvic Fractures and Rectal Injuries

Abstract

Background:

Open pelvic fractures associated with rectal injuries are uncommon. They often cause serious pelvic infection, even death. This combination of injuries has been reviewed infrequently. Herein, we report factors associated with pelvic infection and death in a group of patients with open pelvic fractures and concurrent rectal injuries.

Methods:

We retrospectively reviewed the records of patients with open pelvic fractures and rectal injuries who were treated at our institution from January 2010–April 2014. From the medical records, age, gender, Injury Severity Score (ISS), cause of fracture, associated injuries, classification of the fracture, degree of soft-tissue injury, Glasgow Coma Score (GCS), Revised Trauma Score (RTS), packed red blood cells (PRBCs) needed, presence/absence of shock, early colostomy (yes or no), drainage (yes or no), and rectal washout (yes or no) were extracted. Univariable and multivariable analysis were performed to determine the association between risk factors and pelvic infection or death.

Results:

Twenty patients were identified. Pelvic infection occurred in 50% (n = 10) of the patients. Four patients suffered septicemia, and three patients died of multiple organ dysfunction. The mortality rate thus was 15%. According to the univariable analysis, the patients in whom pelvic infection developed had shock, RTS ≤8, GCS ≤8, blood transfusion ≥10 units in the first 24 h, no colostomy, or Gustilo grade III soft-tissue injury. According to the multivariable analysis, shock and absence of colostomy were independently associated with pelvic infection. By univariable analysis, the only factor associated with death was RTS ≤8.

Conclusion:

The incidence of pelvic infection was lower in patients having early colostomy (p < 0.05). Patients with shock had a higher risk of pelvic infection, and we recommend aggressive measures to treat these patients. According to our results, RTS ≤8 could be a predictor of poor outcomes in patients with open pelvic fracture and concurrent rectal injury. Open reduction and internal fixation after extensive debridement is recommended in patients with unstable pelvic fractures.

Open pelvic fractures caused by high-energy injuries occur more frequently now than in the past as a result of increased road traffic. Open pelvic fractures account for 2%–4% of all pelvic fractures [1 –4]. In the 1970s and 1980s, mortality rates for open pelvic fractures were as high as 50% [5 –7], significantly higher than for closed fractures. With recent developments in medical technology, the mortality rate of open pelvic fracture has declined to 5%–30% [1, 8]. The main cause of early death is critical bleeding, and the main cause of late death is multiple organ dysfunction (MOD) attributable to pelvic infection.

Rectal injuries are a common complication of open pelvic fractures, with an incidence of 17%–64% [1,3,4]. These injuries are associated with heavy tissue contamination leading to infection, and treatment is a challenge, even for experienced surgeons. Dente et al. [9] reported that 55% of patients with open pelvic fractures died early after injury, and that late morbidity and death were associated primarily with pelvic infection. In this paper, we retrospectively analyzed patients with open pelvic fractures and rectal injuries who were treated in our hospital, determining the risk factors in patients with pelvic infection and late death, and we discuss measures that may prevent death.

Patients and Methods

Permission for this study was obtained from the Medical Ethics Committee of Shandong Provincial Hospital Affiliated with Shandong University.

We retrospectively analyzed all patients with open pelvic fractures and concurrent rectal injuries who were treated at our institution from January 2010 to April 2014. We collected the following data for each patient: sex, age, pelvic fracture type, presence/absence of shock, number of associated organ injuries, Glasgow Coma Scale (GCS) score, Injury Severity Score (ISS), Revised Trauma Score (RTS), degree of soft-tissue injury, orthopedic management of the pelvic fracture, colostomy (yes or no), drainage (yes or no), rectal washout (yes or no), days of hospital stay, number of packed red blood cell transfusions in the first 24 h, and death or survival. Several classifications exist for pelvic fractures. We used the Young-Burgess system to classify the bony injury. In brief, it measures anterior–posterior compression injury, lateral compression injury, vertical shear injury, and combined mechanism injury. We used the Gustilo-Anderson system to classify the degree of soft-tissue injury.

Statistical analysis was performed with SPSS Statistics for Windows, Version 12.0 (SPSS Inc., Chicago, IL USA). The mean and standard deviation (SD) were calculated for continuous and discrete variables. All categorical variables were presented as the number of patients and were analyzed using the Fisher exact test. Binary logistic regression analysis was used to identify the risk variables. A two-sided significance value of 0.05 was used for all statistical tests.

Patient characteristics

During the study period, 20 patients met the inclusion criteria. All had full-thickness injuries, with extraperitoneal injury alone identified in 15 patients and both intraperitoneal and extraperitoneal injuries in five. The patients' demographics are shown in in Table 1.

Table 1.

Demographic Information and Trauma Scoring

Characteristic	Average (±SD)	Range
Age	34.7 ± 13.9	16–60
ISS	24.2 ± 15.1	11–75
GCS	11.7 ± 3.4	3–15
RTS	9.2 ± 2.4	6–11
PRBCs	21.1 ± 15.3	0–26

GCS = Glasgow Coma Score; ISS = Injury Severity Score; PRBCs = packed red blood cells; RTS = Revised Trauma Score; SD = standard deviation.

The injuries were severe, with an average ISS of 24.2 ± 15.1 (range 11–75), and 18 patients had an ISS >16. The average blood transfusion requirement in the first 24 h was 21.1 ± 15.3 units (range 0–26 units). The average GCS was 11.7 ± 3.4 (range 3–15), and the average RTS was 9.2 ± 2.4 (range 6–11). The primary mechanism of injury was motor vehicle accident, which accounted for 45% of the patients (n = 9). According to the Young-Burgess classification, the main injury pattern was lateral compression, which was sustained by nine patients (45%). Five patients (25%) suffered anterior–posterior compression injuries, and six patients (30%) had vertical shear injuries.

Shock management and blood transfusion

Eleven patients were in shock at admission and received initial resuscitation. After successful resuscitation, nine patients were transferred to an intensive care unit (ICU), where they remained an average of 16 days (range 5–31 days). Sixteen patients (80%) received a blood transfusion in the first 24 h, and nine patients (45%) were given 10 or more units. The mean blood transfusion requirement within the first 24 h was 21.1 units.

Management of injuries

The orthopedic management of the pelvic fractures was non-operative in 40% of patients (n = 8). Three patients with vertical shear injuries underwent external fixation. Open reduction and internal fixation was performed in eight patients (40%). In two of these patients, the fractures were reduced and fixed when emergency exploratory laparotomy was performed, and fixation was performed in the acute phase in another two patients. The remaining four patients underwent delayed internal fixation. The final patient received a combination of internal and external fixation.

All patients had full-thickness injuries. All patients underwent rectal injury repair; 40% of patients (n = 8) received early colostomy, four patients had delayed colostomy, and seven patients were treated without a colostomy. The details of colostomy surgery are listed in Table 2. Nineteen patients underwent rectal washout with 4–6 L of warm saline, and 13 patients underwent presacral and perianal drainage. All patients received broad-spectrum antibiotics. As soon as possible, we gave antibiotics that could inhibit gram-positive bacteria. Patients who suffer Grade III open fractures should be treated by broad-spectrum antibiotics that inhibit both gram-positive and gram-negative bacteria. If patients had a pelvic infection diagnosed, we gave antibiotics according to the results of a drug sensitivity test.

Table 2.

Number of Patients Treated by Different Methods

	Pelvic Infection		Death
	Y	N	Y	N
Early colostomy	1	7	1	7
Delayed colostomy	3	2	0	5
No colostomy	6	1	2	5

Risk factors for pelvic infection

Pelvic infection occurred in 10 patients (50%), three of whom died of MOD caused by septicemia. Factors associated with pelvic infection by univariable analysis were shock, RTS ≤8, GCS ≤8, blood transfusion ≥10 units in the first 24 h, non-early colostomy, and Gustilo grade III soft-tissue injury. Age, ISS, number of associated injured organs, drainage, rectal washout, Young-Burgess classification of bony fracture, and sex did not influence pelvic infection on univariable analysis.

Using a logistic regression model, we found that shock (odds ratio [OR] 49.772; 95% confidence interval [CI] 6.839–253.610; p = 0.006) and no early colostomy (OR 49.772; 95% CI 6.839–253.610; p = 0.006) were independently associated with pelvic infection.

Risk factors for death

Three patients died (15%), and their mean survival time was 33 days (one died at 10 days, one at 19 days, and one at 70 days). These patients had several common features: Shock, ISS ≥25, RTS ≤8, and blood transfusion requirement ≥10 units in the first 24 h. All of them had pelvic infection, and the main cause of death was pelvic infection leading to MOD. The only factor associated with late death by univariable analysis was RTS ≤8; shock, ISS, and blood transfusion in the first 24 h were not correlated with late death. Details of the univariable analysis for pelvic infection and the overall mortality rate are listed in Table 3.

Table 3.

Risk Factors for Pelvic Infection and Death

	Pelvic infection			Death
	Y	N	P Value	Y	N	P Value
Gender			0.690			0.491
Male	7	8		3	12
Female	3	2		0	5
Age (y)			0.175			0.270
≤30	5	8		1	12
>30	5	2		2	5
Associated injuries			0.375			0.738
None	0	2		0	2
1	1	2		0	3
2	5	4		2	7
≥3	4	2		1	5
Gustilo grade			0.027			0.171
I	0	5		0	5
II	6	4		1	9
III	4	1		2	3
Young class			0.214			0.145
APC	2	5		1	6
LC	4	4		0	8
VS	4	1		2	3
Fracture			0.801			0.344
Nonoperation	4	4		1	7
External fixation	2	1		1	2
Internal fixation^a	4	5		1	8
Shock			0.010			0.193
Absent	1	7		0	8
Present	9	3		3	9
ISS			0.089			0.105
<25	3	7		0	10
≥25	7	3		3	7
GCS			0.043			0.088
≤8	4	0		2	2
>8	6	10		1	15
RTS			0.029			0.031
≤8	6	1		3	4
>8	4	9		0	13
PRBCs (units)			0.035			0.074
≥10	7	2		3	6
<10	3	8		0	11
Early colostomy			0.010			0.656
No	9	3		2	10
Yes	1	7		1	7
Drain			0.5			0.270
Absent	3	4		2	5
Present	7	6		1	12
Rectal washout			0.5			0.150
No	1	0		1	0
Yes	9	10		2	17

Patients treated with internal and external fixation.

APC = anterior–posterior compression; GCS = Glasgow Coma Score; ISS = Injury Severity Score; LC = lateral compression; PRBCs = packed red blood cells; RTS = Revised Trauma Score; VS = vertical shear.

Discussion

During the study period, pelvic fractures accounted for approximately 2% of all orthopedic admissions to our hospital. Pelvic fracture trauma is severe; the associated conditions are complex, and the mortality rate is high. The reported mortality rate caused by open pelvic fractures is 5%–20% [1, 8]. Despite recent advances in the management of open pelvic fractures, the treatment of these injuries remains an interesting challenge. The fracture fragments can puncture vaginal, rectal, and perineal tissues and the skin. Chong et al. [10] indicated that one-third of patients died from associated complications. Rectal injury leads to bacterial contamination from the intestine and increases the risk of pelvic infection. Kataoka et al. [11] reported that late deaths in patients with open pelvic fractures and concurrent rectal injuries almost invariably result from sepsis.

In general, pelvic fracture is a high-energy injury, causing special concern about instability, which is easily complicated by hemorrhagic shock associated with damage to the unique anatomic structures. The pelvis consists mainly of cancellous bone, and many kinds of soft tissue are in and around the site, especially the numerous arterial plexus and venous vessels. Therefore, the pelvis has a rich blood supply. The space between the pelvic cavity and the posterior peritoneum consists largely of loose connective tissue that can accommodate enormous amounts of blood. The blood loss can be as high as 3 L. Meanwhile, rectal injuries allow a large number of bacteria to translocate into the pelvic cavity. As is well known, blood is an excellent culture medium, which promotes bacterial proliferation and leads to pelvic infection. In the current study, shock frequently was associated with pelvic infection, so effective anti-shock therapy is an important part of acute treatment and could decrease the rate of pelvic infection and save a life [13].

Early blood transfusion is the most important method of shock therapy. Brenneman et al. [2] reported that patients with open pelvic fractures often require more blood than those with closed pelvic fractures, with the former receiving blood transfusions in the first 24 h at a four times greater rate than the latter. Dente et al. [9] reviewed 44 patients with open pelvic fractures and reported that the mean transfusion requirement was 11.5 units during the first 24 h. In our study, the mean blood transfusion requirement in the initial resuscitation phase was 21.1 units. Additionally, temporarily stabilizing the pelvis and controlling pelvic volume are important, and can be achieved with pelvic binders, C-clamps, or external fixators. In our study, pelvic binders were used in two patients. Pelvic fractures often carry a higher risk when combined with pelvic arterial injury, and patients with uncontrolled hemorrhage should receive diagnostic angiography [14, 15]; however, this technique is seldom used in our institution.

Control of hemorrhage is fundamental; therefore, emergency stabilization of the pelvis is essential followed by surgical procedures [16]. However, the ideal fixation is controversial. Generally, external fixation is the first choice for open pelvic fractures to avoid pelvic infection. Chen et al. [17] found that internal fixation was a better choice in the treatment of open pelvic fractures. Nine patients underwent open reduction and internal fixation, and four patients developed pelvic infection. Statistical analysis showed that there was no significant difference between the internal and external fixation groups. Malunion and dysfunction occur often in patients undergoing external fixation and are associated with unstable pelvic fractures because of inappropriate reduction or absence of complete reduction. One study indicated that patients with unstable pelvic fractures should undergo open reduction and internal fixation after extensive cleaning and lavage, which could reduce the mortality rate and achieve good functional results [18, 19]. Also, patients requiring exploratory laparotomy should receive internal fixation to avoid a second operation.

Hasankhani et al. [20] studied 15 patients with open pelvic fractures associated with extensive perineal injuries and concluded that early colostomy improves outcomes and reduces the mortality rate. However, others found no difference in the overall infectious complication rate between the colostomy and no-colostomy groups [21,22]. Our results support the conclusions of Hasankhani et al. Eight of our patients underwent early colostomy and 12 patients did not; 9 of the 12 patients without early colostomy developed pelvic infection, and two of three patients without early colostomy died of MOD resulting from sepsis. Statistical analysis showed that the overall infectious complication rate was significantly different in the early colostomy and non-early colostomy groups and that there was no statistically significant difference between the drained and non-drained groups or between the washout and non-washout groups regarding pelvic infection. In theory, fecal diversion can prevent further contamination [23], presacral and perianal drainage can prevent the development of closed-space abscess [24], and rectal washout can decrease the chance of bacterial translocation [25]. Therefore, we believe that presacral and perianal drainage and rectal washout are necessary in patients with rectal injury even if these measures have no significant influence on the pelvic infection rate.

In our patients, the overall pelvic infection rate was associated with shock, RTS, GCS, blood transfusion in the first 24 h, colostomy, and Gustilo classification of soft-tissue injury. Also, our multivariable analysis showed that shock and absence of early colostomy were independently associated with pelvic infection. Jones et al. [26] reviewed 16 patients with open pelvic fractures and rectal injuries and reported a pelvic infection rate of 70%. This article also found that early colostomy could decrease the pelvic infection rate [26]. We agree with these conclusions and constructed a treatment scheme for these patients: (1) hemorrhage control and resuscitation; (2) extensive debridement; (3) early colostomy; (4) drainage and rectal washout; and (5) use of antibiotics.

It has been reported that death correlates with RTS, ISS, Gustilo grade III soft-tissue injury, GCS, and vertical shear bony injury. Late death was associated with ISS, RTS, pelvic sepsis, Gustilo classification of soft-tissue injury, and blood transfusion in the first 24 h [18]. In our patients, the mortality rate was 15%, and all deaths were late and associated with RTS. The RTS includes the GCS, systolic blood pressure, and ventilatory rate. The score is highly sensitive and can assess the degree of trauma simply [27]. The mortality rate in our patients with an RTS ≤8 was high. Based on our experience, maintaining airway patency, effective gas exchange, supplementing circulating blood volume, early colostomy, and fixation of pelvic fractures are necessary, and these measures decrease the mortality rate.

The first limitation of our study is its retrospective design. Another limitation is that we excluded patients with rectal mucosal laceration to gain more accurate statistical results. The third limitation is that our study was conducted on a small number of patients because of the low morbidity.

Conclusion

Open pelvic fractures associated with rectal injury have a high mortality rate, and these injuries pose a challenge to even experienced surgeons. Our experience treating these patients revealed that the main cause of late death was MOD as a result of pelvic infection, which was associated with shock, RTS, GCS, blood transfusion in the first 24 h, non-early colostomy, and Gustilo classification of soft-tissue injury. Early colostomy could significantly decrease the pelvic infection rate. Patients with shock have a high risk of pelvic infection, and we recommend aggressive measures to treat these patients. Based on our results, RTS ≤8 can be used to predict poor outcomes in open pelvic fractures with concurrent rectal injury. Also, open reduction and internal fixation after extensive cleansing and lavage is an option in patients with unstable pelvic fractures.

Footnotes

Author Disclosure Statement

All authors warrant that they have no conflict of interest in connection with this article. They all had access to all the data used in the study and hold final responsibility for the decision to submit this manuscript for publication.

This study was supported by the Key R&D program in Shandong Province, China (No. 2016GSF201214) and the National Natural Science Foundation of China (No. 81672156).

References

Davidson

, Simmons

, Williamson

, Buerk

. Pelvic fractures associated with open perineal wounds: A survivable injury. J Trauma, 1993; 35:36–39.

Brenneman

, Katyal

, Boulanger

, et al. Long-term outcomes in open pelvic fractures. J Trauma, 1997; 42:773–777.

Ferrera

, Hill

. Good outcomes of open pelvic fractures. Injury, 1999; 30:187–190.

Hanson

, Milne

, Chapman

. Open fractures of the pelvis: Review of 43 cases. J Bone Joint Surg, 1991; 73B:325–329.

Perry

Jr.

Pelvic open fractures. Clin Orthop Rel Res, 1980:41–45.

Rothenberger

, Fischer

, Strate

, et al. The mortality associated with pelvic fractures. Surgery, 1978; 84:356–361.

Maull

, Sachatello

, Ernst

. The deep perineal laceration—An injury frequently associated with open pelvic fractures: A need for aggressive surgical management: A report of 12 cases and review of the literature. J Trauma, 1977; 17:685–696.

Richardson

, Harty

, Amin

, Flint

. Open pelvic fractures. J Trauma, 1982; 22:533–538.

Dente

, Feliciano

, Rozycki

, et al. The outcome of open pelvic fractures in the modern era. Am J Surg, 2005; 190:830–835.

10.

Chong

, DeCoster

, Osler

, Robinson

. Pelvic fractures and mortality. Iowa Orthop J, 1997; 17:110–114.

11.

Kataoka

, Minehara

, Shimada

, et al. Sepsis caused by peripelvic soft tissue infections in critically injured patients with multiple injuries and unstable pelvic fracture. J Trauma, 2009; 66:1548–1554.

12.

Routt

Jr , Simonian

, Ballmer

. A rational approach to pelvic trauma: Resuscitation and early definitive stabilization. Clin Orthop Rel Res, 1995;(318):61–74.

13.

Grotz

, Allami

, Harwood

, et al. Open pelvic fractures: Epidemiology, current concepts of management and outcome. Injury, 2005; 36:1–13.

14.

Costantini

, Coimbra

, Holcomb

, et al. Current management of hemorrhage from severe pelvic fractures: Results of an American Association for the Surgery of Trauma multi-institutional trial. J Trauma Acute Care Surg, 2016; 80:717–723.

15.

Perez

, Hughes

, Bowers

. Angiographic embolisation in pelvic fracture. Injury, 1998; 29:187–191.

16.

Westhoff

, Holl

, Kalicke

, et al. [Open pelvic fracture: Treatment strategy and results for 12 patients] (Ger). Unfallchirurg, 2004; 107:189–196.

17.

Chen

, Zhang

, Wu

, et al. Percutaneous limited internal fixation combined with external fixation to treat open pelvic fractures concomitant with perineal lacerations. Orthopedics, 2011; 34:e827–e831.

18.

Dong

, Zhou

. Management and outcome of open pelvic fractures: A retrospective study of 41 cases. Injury, 2011; 42:1003–1007.

19.

Leenen

, van der Werken

, Schoots

, Goris

. Internal fixation of open unstable pelvic fractures. J Trauma, 1993; 35:220–225.

20.

Hasankhani

, Omidi-Kashani

. Treatment outcomes of open pelvic fractures associated with extensive perineal injuries. Clin Orthop Surg, 2013; 5:263–268.

21.

Lunsjo

, Abu-Zidan

. Does colostomy prevent infection in open blunt pelvic fractures? A systematic review. J Trauma, 2006; 60:1145–1148.

22.

Kottmeier

, Wilson

, Born

, et al. Surgical management of soft tissue lesions associated with pelvic ring injury. Clin Orthop Rel Res, 1996;(329):46–53.

23.

Grasberger

, Hirsch

. Rectal trauma: A retrospective analysis and guidelines for therapy. Am J Surg, 1983; 145:795–799.

24.

Ganchrow

, Lavenson

Jr , McNamara

. Surgical management of traumatic injuries of the colon and rectum. Arch Surg, 1970; 100:515–520.

25.

Moore

, Jones

. Benefits of immediate jejunostomy feeding after major abdominal trauma: A prospective, randomized study. J Trauma, 1986; 26:874–881.

26.

Jones

, Powell

, Kellam

, et al. Open pelvic fractures: A multicenter retrospective analysis. Orthop Clin North Am, 1997; 28:345–350.

27.

Champion

, Sacco

, Copes

, et al. A revision of the Trauma Score. J Trauma, 1989; 29:623–629.