The impact of mechanism on the management and outcome of penetrating colonic trauma

Abstract

Introduction

In light of continuing controversy surrounding the management of penetrating colonic injuries, we set out to compare the outcome of penetrating colonic trauma according to whether the mechanism of injury was a stab wound or a gunshot wound.

Methods

Our trauma registry was interrogated for the 5-year period from January 2012 to December 2016. All patients over the age of 18 years with penetrating trauma (stab or gunshot) and with intraoperatively proven colonic injury were reviewed. Details of the colonic and concurrent abdominal injuries were recorded, together with the operative management strategy. In-hospital morbidities were divided into colon-related and non-colon related morbidities. The length of hospital stay and mortality were recorded. Direct comparison was made between patients with stab wounds and gunshot wounds to the colon.

Results

During the 5-year study period, 257 patients sustained a colonic injury secondary to penetrating trauma; 95% (244/257) were male and the mean age was 30 years. A total of 113 (44%) sustained a gunshot wound and the remaining 56% (144/257) sustained a stab wound. Some 88% (226/257) of all patients sustained a single colonic injury, while 12% (31/257) sustained more than one colonic injury. A total of 294 colonic injuries were found at laparotomy. Multiple colonic injuries were less commonly encountered in stab wounds (6%, 9/144 vs. 19%, 22/113, P < 0.001). Primary repair was more commonly performed for stab wounds compared with gunshot wounds (118/144 vs. 59/113, P < 0.001). Patients with gunshot wounds were more likely to need admission to intensive care, more likely to experience anastomotic failure, and had higher mortality.

Conclusions

It would appear that colonic stab wounds and colonic gunshot wounds are different in terms of severity of the injury and in terms of outcome. While primary repair is almost always applicable to the management of colonic stab wounds, the same cannot be said for colonic gunshot wounds. The management of colonic gunshot wounds should be examined separately from that of stab wounds.

Introduction

The management of penetrating colonic trauma is controversial and has evolved over the past 50 years.^1–10 Historically, civilian colonic trauma was managed with a diverting colostomy. Primary repair was actively discouraged because of the perceived high risk of anastomotic breakdown. This was as a direct result of surgical experience during the Second World War, where the introduction of mandatory diverting colostomy dramatically reduced the morbidity and mortality associated with colonic injuries.^1,2 This approach held sway for three decades following the War and was only challenged in the late 1970s.^3–15 As resuscitation, operative techniques, anaesthesia and prehospital retrieval all improved, a substantial body of literature, predominantly from South Africa and the United States, emerged to show that primary repair of most colonic injuries is safe.^3,10,13 The previously accepted contraindications to primary repair such as shock, peritonitis, concurrent renal injury and delay were increasingly questioned. By the turn of the millennium, the Eastern Association for the Surgery of Trauma (EAST) guidelines suggested that there was sufficient high-quality evidence to support a standard of primary repair for all non-destructive colonic injuries.¹⁶ Since the turn of the millennium, the concept of damage control surgery and damage control resuscitation have become firmly established in trauma care.^17–20 This has tended to bring the debate about the management of colonic trauma full circle, as complex definitive repairs are now eschewed in favour of temporising and abbreviated techniques that allow for the patient to be taken to intensive care for physiological stabilisation before returning to the operating room for definitive management of the intra-abdominal injuries. In conjunction with these changes, the nature of urban violence and crime has continued to change. Gun crime is a serious problem in inner city areas of the US and in the peri-urban townships in South Africa. In addition, the world is experiencing a prolonged period of financial and political instability. Even countries which have enjoyed a protracted period of peace and prosperity for the last 70 years are now confronted by the spectre of urban terrorism. Both knife- and gun-related trauma are likely to be features of the urban environment for the foreseeable future. Our centre has always managed a significant burden of trauma and 5 years ago introduced a hybrid electronic medical registry (HEMR), which has massively facilitated data collection. This has allowed us to document our not inconsiderable experience. In light of this experience, we set out to compare penetrating colonic trauma according to whether the mechanism was a stab or a gunshot.²¹

Materials and Methods

Clinical setting

This retrospective study was undertaken at the Pietermaritzburg Metropolitan Trauma Service (PMTS), Pietermaritzburg, South Africa. The HEMR was reviewed for the 5-year period from January 2012 to December 2016. Ethics approval for this study and for maintenance of the registry was obtained from the Biomedical Research Ethics Committee of the University of Kwa Zulu Natal (reference: BE 207/09 and BCA 221/13). The PMTS provides definitive trauma care to the city of Pietermaritzburg, the capital of Kwa Zulu Natal province. It is one of the largest academic trauma centres in Kwa Zulu Natal and also serves as the referral centre for 19 rural hospitals within the province, with a total catchment population of over three million. Over 50% of patients managed by the PMTS sustain penetrating trauma.

The study

All patients over the age of 18 years who were subjected to laparotomy for penetrating trauma (gunshot or stab wounds) and who sustained an intraoperatively proven colonic injury were included in this study. Basic demographic data, mechanism of injury and admission physiology were reviewed, together with the operative records for each patient. Details of the site of the colon injury, the operative management strategy and any concurrent abdominal injuries noted at laparotomy were recorded. All in-hospital morbidities were reviewed. Morbidity was divided into colon- and non-colon-related morbidity. All anastomotic leaks, intra-abdominal sepsis and wound sepsis were regarded as colonic-related complications. All other septic complications such as pneumonia or urinary tract infection were regarded as non-colonic-related complications. The length of hospital stay and mortality were recorded. A direct comparison was made between patients with gunshot wounds and stab wounds of the colon.

Management

Our standard approach to the injured colon is to debride and primarily repair all perforations with a single layer absorbable suture. A nondestructive injury is defined as a wound involving less than 50% of the bowel wall without devascularisation. A diverting colostomy is performed only if the colonic damage was so extensive that a resection was deemed necessary and there was disseminated gross faecal contamination or pus. Damage control surgery was indicated in the face of hypothermia less than 34°C, acidosis as defined by a pH less than 7.2 or a revised trauma score less than 5. If damage control was adopted, diversion was not done but the bowel stapled off and placed back into the abdomen. Once the physiological homeostasis had been achieved, the patient was returned to theatre for definitive management. The management of the colonic injury in this setting was left at the discretion of the individual surgeon’s preference.

Statistical analysis

Data were processed and analysed using Stata 13.0 (StataCorp. 2013. Stata Statistical Software: Release 13. StataCorp LP, College Station, TX). Continuous variables were summarised using mean and standard deviation (SD). If there was evidence of skewing or asymmetry, median and interquartile range (IQR) were presented instead. Differences in means of continuous variables such as age were assessed using the Student’s t-test. Association between categorical variables were assessed using the Pearson chi-squared test or the Fisher’s exact test if expected cell count with fewer than observations. The significance cut-off used was 5%.

Results

Overview

During the 5-year study period, 257 patients sustained a colonic injury secondary to penetrating trauma; 95% (244/257) were male and the mean age was 30 years. A total of 113 (44%) sustained a gunshot wound and the remaining 56% (144/257) sustained a stab wound. The admission physiology was as follows: heart rate: 103 beats/minute, systolic blood pressure: 103 mmHg, pH 7.35, lactate 3.1 mmol/l.

Operative findings

A total of 85% (226/257) of all patients sustained a single colonic injury, while 12% (31/257) sustained more than one colonic injury. A total of 294 colonic injuries were found at laparotomy. Table 1 summarises the site of colon injuries.

Table 1.

Colon injury site

Injury site	Patients (N = 294)
Caecum	30
Ascending colon	20
Hepatic flexure	17
Transverse colon	102
Splenic flexure	28
Descending colon	52
Sigmoid colon	45

Operative strategies

There were 177 colon injuries in which primary repair was undertaken. Colostomy was performed in 31 cases. Resection and primary anastomosis was performed in 26 cases. In 23 cases, damage control surgery was performed. In this cohort of patients, the injured bowel was stapled off as part of damage control strategy and left inside the abdomen. Table 2 summarises the concurrent non-colonic intra-abdominal injuries.

Table 2.

Concurrent non-colonic intra-abdominal injuries

Concurrent injuries	Patients (N = 338)
Small bowel	101
Diaphragm	47
Stomach	46
Liver	39
Kidney	28
Spleen	19
Pancreas	16
Intra-abdominal vessels	16
Duodenum	10
Mesentery	10
Bladder	3
Gallbladder	3

Clinical outcomes

Some 34% (88/257) of patients required intensive care admission. Of the 257 patients, 67 (25%) experienced one or more complications; 3 of the 177 (1.7%) patients who underwent primary repair of a colonic injury and 4 of 26 patients (15%) who underwent resection and primary anastomosis experienced an anastomotic leak and developed intra-abdominal sepsis. All seven of these patients underwent repeat laparotomy. All were defunctioned at repeat laparotomy. One of this group died from multiple organ dysfunction. The mean length of stay for all patients was 12 days. The overall mortality was 10% (25/257).

Stab wounds compared with gunshot wounds

Direct comparison was made between those patients who sustained a stab wound and those who sustained a gunshot wound. Table 3 summarises this comparison. Multiple colonic injuries were less commonly encountered in stab wounds (6%, 9/144, vs. 19%, 22/113, P < 0.001). Primary repair was more commonly performed in stab wounds compared with gunshot wounds (118/144 vs. 59/113, P < 0.001). Mortality was similar in both groups. Patients with gunshot wounds were more likely to need admission to intensive care, more likely to experience anastomotic failure , and had higher mortality.

Table 3.

Stab wounds compared with gunshot wounds in colon trauma

Characteristics	Stab (N = 144)	Gunshot (N = 113)	P value
Demographics:
Age (years)	29	31	0.259
Male n (%)	135 (94)	109 (96)	0.396
Female n (%)	9 (6)	4 (4)
Admission physiology:
Mean heart rate (beats/minute)	98	103	0.151
Mean systolic pressure (mmHg)	118	121	0.518
Mean pH	7.37	7.35	0.371
Mean lactate level (mmol/l)	2.7	3.1	0.356
Injury pattern:
Single n (%)	135 (94)	91 (81)	0.001
Multiple n (%)	9 (6)	22 (19)
Colon injury site:
Caecum n (%)	12	18	0.047
Ascending colon n (%)	5	15	0.005
Hepatic flexure n (%)	6	11	0.166
Transverse colon n (%)	53	49	0.289
Splenic flexure n (%)	18	10	0.403
Descending colon n (%)	35	17	0.051
Sigmoid colon n (%)	24	21	0.495
Operative strategies:
Resection n (%)	20	32	0.016
Ligation n (%)	8	22	0.007
Primary repair n (%)	118	59	< 0.001
Colostomy n (%)	15	16	0.559
Anastomosis n (%)	6	20	0.001
Leak n (%)	1	6	0.024
Clinical outcomes n (%)	67 (47)	67 (59)	0.501
Intensive care admission n (%)	36 (25)	52 (46)	0.002
Mean length of stay (days)	12	12	0.495
Mortality n (%)	7 (5)	18 (16)	0.014

Discussion

At the conclusion of the Second World War, military surgeons returned to civilian practice and propagated the dogma of universal faecal diversion that they had learnt on the battlefield.^1–3 It was only in 1979 that Stone and Fabian initiated a paradigm shift in the management of these injuries when they published a prospective, non-blinded study in which 139 patients were randomised to either primary repair or faecal diversion.⁴ The authors found equivalent rates of infection (48% vs. 57%, P > 0.05) and mortality (1.5% vs. 1.4%, P > 0.05) between the two groups. However, they excluded 129 patients who underwent an obligatory colostomy due to the presence of either shock, blood loss greater than 1000 ml, more than two intra-abdominal organs injured, significant peritoneal contamination, delay greater than 8 hours after injury, destructive colon wounds or major abdominal wall loss.

This pioneering study was followed by many retrospective cohort type reports with varying numbers of patients.^5,6 The results remained consistent in that primary repair was as efficacious as diversion and that the fear of anastomotic failure and intra-abdominal leakage was overstated. The next well-designed prospective trial was published in 1991 and focused on a cohort of 56 patients with penetrating colon injuries randomised to either primary repair or diversion.⁷ These authors did not advocate an ‘obligatory colostomy’ for any subset of patients. Once again, they showed that septic complications were similar between the two groups and there were no anastomotic leaks observed in the primary repair cohort. While US surgeons were increasingly advocating primary repair for colonic injury, surgeons in South Africa such as Demetriades at Baragwanath Hospital in Johannesburg and Baker at King Edward VIII Hospital in Durban also increasingly reported on the successful application of primary repair of penetrating colonic injuries.^8–10 Following a number of prospective and retrospective studies the Eastern Association for the Surgery of Trauma practice management guidelines advocated that only patients with destructive colon wounds with a penetrating abdominal trauma index greater than 25, significant comorbidities or haemodynamic instability should undergo diversion.^11–16

Our current data are similar to those reported over three decades ago by Demetriades et al.¹⁰ At the time, he and his colleagues performed primary repair for almost all gunshot injuries to the colon. However, they still used primary diversion if ‘the colonic damage was so extensive that a resection was deemed necessary, and there was disseminated gross faecal contamination or pus’. Of a total of 102 patients, primary repair was performed in 76%, a Hartmann's procedure in 16% and repair of the wound with a proximal colostomy in 8%. Although Demetriades was adamant that gunshot and stab wounds behaved in a similar fashion, his own data suggest that this is only partly true. It is apparent from the literature that most authors, even those most supportive of primary repair, consider that certain situations demand a diversion. These factors have not changed much since Stone’s seminal paper in 1979 and include the presence of shock, faecal contamination, the severity of the colonic injury itself and the number of concomitant organ injuries. Our data suggest that it is the gunshot cohort that is most likely to exhibit these features. It would therefore appear that management strategies will differ depending on the mechanism of the penetrating trauma. Although most of the civilian gunshot wounds encountered are low velocity type wounds, it would appear that they are more destructive than simple stab wounds. In our series, patients with gunshot wounds were more likely to have associated multiple colonic injuries and were more likely to require damage control ligation, resection and anastomosis. They were also more likely to not undergo a primary repair than a patient with a stab wound and much more likely to develop complications and an anastomotic leak. Patients with gunshot wounds were significantly more likely to require intensive care admission and have a much higher mortality rate than those with stab wounds.

Since the turn of the millennium, the concepts of damage control surgery and resuscitation have become firmly established.^17–20 Complex definitive repairs are now eschewed in favour of temporising and abbreviated techniques which allow for the patient to be taken to intensive care for physiological stabilisation before returning to the operating room for definitive management of the intra-abdominal injuries. This has brought the debate full circle.²¹ The cohort of patients requiring an abbreviated laparotomy will meet most of the traditional criteria for diversion. They will be shocked, acidotic and hypothermic and will most likely have concomitant injuries and faecal contamination, as well as destructive colonic injuries. It would seem that, in light of this, the role of primary diversion should be minimal. Either the patient is not physiologically deranged and does not require a damage control operation or the patient’s physiological derangement necessitates an abbreviated operation. In the first instance, primary repair, including resection and anastomosis, is appropriate. If the latter situation prevails then primary diversion is contraindicated. In this situation, the bowel ends must be stapled off and placed in the abdomen to await definitive management once the patient’s physiology has been restored. The debate should now increasingly shift towards the nature of this definitive management. Should the patient undergo a delayed primary repair or a delayed stoma at the repeat surgery is the question which now needs to be answered.

Conclusion

It would appear that colonic stab wounds and colonic gunshot wounds are different in terms of severity of the injury and in terms of outcome. This means that while primary repair is almost always applicable to the management of colonic stab wounds the same cannot be said for colonic gunshot wounds. The management of colonic gunshot wounds should be examined separately from that of stab wounds.