Civilian rectal trauma: a changing perspective

Civilian rectal trauma: A changing perspective Jeffrey J. Morken, MD, James J. Kraatz, MD, Emmanuel G. Balcos, MD, Mark J. Hill, MD, Arthur L. Ney, MD, Michael A. West, MD, Joan M. Van Camp, MD, Richard T. Zera, MD, Donald M. Jacobs, MD, Mark D. Odland, MD, and Jorge L. Rodriguez, MD, Minneapolis, Minn

Background. Recently the Organ Injury Scaling Committee of the American Association for the Surgery of Trauma developed a Rectal Injury Scaling System (RISS). Little data exist regarding its clinical utility. Methods. We retrospectively reviewed 45 patients with rectal injuries to assess the impact of the RISS on patient management and outcome. We compared RISS grade I patients (group I, partial-thickness injury) with patients with grades 2, 3, and 4 injuries (group II, full-thickness injury). Results. Group II underwent distal rectal washout and repair of the injury twice as often and had a significantly higher rate of diversion of the fecal stream. This was associated with a 3-fold increase in complications. The only complications in group I were in patients managed with diversion of the fecal stream and distal rectal washout. Conclusions. Our data suggest that aggressive surgical management for RISS grade I injury may not be necessary. Implementation of therapy based on the RISS may improve outcomes of civilian rectal trauma. (Surgery 1999;126:693-700.) From the Department of Surgery, Division of Trauma, Hennepin County Medical Center, Minneapolis, Minn

THE MANAGEMENT PRINCIPLES OF civilian rectal injuries have evolved from lessons learned during wartime experiences1-6 including diversion of the fecal stream, debridement and closure of the rectal injury when possible, distal rectal washout, placement of a presacral drain, and perioperative broad-spectrum antibiotics. Adhering to such dogma has produced a low mortality from civilian rectal injuries, but morbidity is still significant.7-12 Recently, several mainstays of the operative management have been challenged with no increase in mortality, specifically diversion of the fecal stream,13 distal rectal washout,8,11,14 and placement of a presacral drain.10,14,15 Unfortunately, these reports fail to categorize the severity of the rectal injury. This lack of specific injury data leaves surgeons dealing with such injuries in a quandary regarding the utilization of surgical concepts that in the past have lead to improved outcomes. In the late 1980s the Organ Injury Scaling Committee of the American Association for the Surgery of Trauma (OISC-AAST) was charged to devise injury severity scores for individual organs. Presented at the 56th Annual Meeting of the Central Surgical Association, St Louis, Mo, Mar 4-6, 1999. Reprint requests: Jorge L. Rodriguez, MD, Department of Surgery, Hennepin County Medical Center, 701 Park Ave, Minneapolis, MN 55415. Copyright © 1999 by Mosby, Inc. 0039-6060/99/$8.00 + 0

11/6/99882

In 1990 the second report from this committee proposed the Rectal Injury Scaling System (RISS),16 a graded anatomic description scaled from 1 to 5. Grade 1 represents the least severe injury, a partial-thickness injury of the rectum. Grades 2, 3, and 4 represent graded full-thickness injuries in an otherwise intact, vascularized rectum. Grade 5, the most severe injury, represents a devascularized segment of rectum. To date, little attention has been directed toward the RISS as it relates to operative management of these injuries. Severity of the rectal injury as an independent variable has not been considered in most of the literature, because all rectal injuries are grouped together. The possibility that the operative management of these injuries can be very different depending on the severity of the rectal injury has not been clearly examined. We therefore undertook this review to assess the possible impact of the RISS on operative management and outcome of civilian rectal injuries. METHODS We performed a retrospective medical record and trauma registry review of all patients seen and treated for rectal trauma at Hennepin County Medical Center, an urban Level 1 Trauma Center. To be included in the review, all patients had to have an extraperitoneal rectal injury incurred during blunt or penetrating trauma in unprepared large intestine diagnosed by physical examination SURGERY 693

694 Morken et al

Surgery October 1999

Table I. Total group (N = 45) demographic data Age (y) Gender (% male) Mechanism of injury (% penetrating) Injury severity score Number of associated injuries Admission Glasgow Coma Score Systolic blood pressure < 90 mm Hg on admission (%)

29 ± 14 82 84 21 ± 12 4±3 14 ± 1 25

or proctosigmoidoscopy. Data collected included demographics, management, and outcomes. The RISS was calculated based on the description of the injury on physical examination or proctosigmoidoscopy. For comparison, patients were categorized into 2 groups. Group I was comprised of patients with a partial-thickness injury of the rectum—RISS grade 1. Group II included patients with a full-thickness injury in an otherwise intact, vascularized rectum—RISS grades 2, 3, and 4. In addition, a subgroup analysis was performed in group I patients to assess the impact of diversion of the fecal stream. Nonparametric data were assessed using Mann-Whitney U analysis and are reported as mean ± standard deviation. Parametric data were assessed using Fisher exact test and are presented as a percentage. A P value of less than .05 was considered significant. RESULTS From May 1988 through September 1998, 45 patients were treated for rectal trauma that met inclusion criteria. These patients were young men injured by penetrating trauma that inflicted multiple injuries not associated with significant hemodynamic alteration (Table I). Physical examination, either preoperative or intraoperative, and proctosigmoidoscopy were used to diagnose the injury, with gross blood or laceration being the most common findings. Most of the injuries were full thickness. Distal rectal washout, repair of the injury, complete diversion of the fecal stream within 100 ± 175 minutes from injury, and perioperative antibiotics were the mainstays of management. Only a small number of patients had placement of a presacral drain (Table II). These patients used considerable hospital resources, with a third having significant morbidity. Less than half of the patients undergoing colostomy for diversion of the fecal stream had their colostomy reversed (Table III). Twenty-nine percent (13 of 45) of patients had a RISS grade 1 injury (partial thickness) and these patients comprised group I. Seventy-one percent

Table II. Total group (N = 45) injury and management data Digital rectal examination (%) Performed Positive finding Gross blood Laceration Foreign body Proctosigmoidoscopy (%) Performed Positive finding Hematoma Gross blood Laceration RISS (%) Grade I Grade II Grade III Grade IV Grade V Surgical management (%) Distal rectal washout Damage repair Diversion of the fecal stream Drainage of presacral space Perioperative antibiotics

89 68 15 70 15 58 100 15 15 70 29 42 11 18 0 53 46 78 26 100

(32 of 45) had a RISS grade 2, 3, or 4 injury (full thickness) and comprised group II. No patient had a RISS grade 5 injury. There was no difference in age, gender, mechanism of injury, admission Glasgow Coma Score, admission hypotension (systolic blood pressure less than 90 mm Hg), or time to definitive care between the 2 groups. Group II had a statistically significant greater injury severity score and number of associated injuries (Table IV), with the most common associated injuries including 11 pelvic fractures, 10 abdominal and pelvic vascular injuries, 9 small bowel injuries, 9 soft tissue injuries, and 6 bladder injuries. In contrast, in group I the most common injuries included 2 extremity fractures, 2 vascular injuries, 1 pelvic fracture, 1 bowel injury, and 1 bladder injury. All injuries were diagnosed by either physical examination (preoperative or intraoperative) or by sigmoidoscopy. Digital rectal examination was used in 92% of patients in group I and 88% in group II, with 75% and 64% being positive, respectively, for either gross blood or laceration. Sixty-nine percent of patients in group I and 72% in group II had a stool Hemoccult test performed with a 90% and 73% positive rate, respectively. Sixty-nine percent of patients in group I and 53% in group II underwent sigmoidoscopy, of which 100% and 94% were

Morken et al 695

Surgery Volume 126, Number 4 Table III. Total group (N = 45) outcome data Hospital length of stay (days) Mortality (%) Morbidity (%) Colostomy closed (%) Time of colostomy closure (mo)

Table IV. Demographic data for group I vs group II Parameters Age (y) Gender (% male) Mechanism of injury (% penetrating) Injury severity score Number of associated injuries Admission Glasgow Coma Score Systolic blood pressure < 90 mm Hg on admission (%) Time to definitive care (min)

17 ± 20 4 33 49 3.5 ± 2.6

Group I (n = 13)

Group II (n = 32)

30 ± 10 77 92 14 ± 9 2±2 14 ± 0.4 15 99 ± 102

29 ± 15 84 76 23 ± 13* 4 ± 3* 14 ± 1 30 101 ± 192

*P = < .05.

positive, respectively, for either gross blood or laceration. All group II patients had an exploratory celiotomy. Twenty-six patients were assessed for abdominal injury: 16 gunshot wounds (9 positive), 5 impalement injuries (1 positive), 2 crush injuries to the pelvis (2 positive), 1 bleeding from a perinealinguinal laceration (1 positive), 1 abdominal computed tomographic scan suspicious for a colonic injury (1 negative), and 1 vehicular accident (1 positive). Six had a celiotomy for management of the rectal injury: 6 vehicular accidents (6 negative). In contrast, 62% (8 of 13) of group I patients underwent an exploratory celiotomy. Seven patients were assessed for abdominal injury: 7 gunshot wounds (5 positive). One had a celiotomy for management of the rectal injury: 1 motor vehicle accident (1 negative). Group II underwent distal rectal washout and repair of the injury twice as often as group I and had a significantly higher rate of diversion of the fecal stream. The method of diversion was equally distributed between loop and end colostomy. Placement of a presacral drain was less common in group II (Table V). Group II had a significant 3-fold increase in complications, which was associated with a significant use of hospital resources. In group II, 8 patients had a midline abdominal wound infection with cellulitis that required opening of the wound and intravenous antibiotics. Two patients had bleeding from the rectal stump that required operative intervention. Two patients developed a pelvic abscess that required drainage and intravenous antibiotics, and 1 patient dehisced and eviscerated

his wound, requiring operative intervention. In group I, 2 patients had complications, 1 patient had a dehiscence of his wound that was treated nonoperatively and the other had a dehiscence with evisceration that required operative intervention. Mortality was greater in group II but this difference was not significant. One patient in group II died within 24 hours of injury as a result of severe hemorrhage and the other as a result of sepsis from mucormycosis 6 days after injury initially diagnosed in an extremity wound. Fewer patients had their colostomy reversed in group II and the timing from injury to colostomy reversal was significantly longer (Table VI). The subgroup analysis of group I was performed between patients who underwent fecal diversion (n = 6) and those who did not (n = 7) to assess the impact of this procedure. There was no difference in age, mechanism of injury, admission Glasgow Coma Score, admission hypotension, injury severity score, or time to definitive care between the two subgroups. Patients who had their fecal stream diverted had a significant number of associated injuries (Table VII). Of the patients in whom the fecal stream was diverted, 5 (85%) had distal rectal washout, 1 had the rectal injury repaired (17%), and 4 (67%) had placement of a presacral drain. Patients who were not diverted had neither distal rectal washout nor did they have a presacral drain placed and 2 patients (29%) had primary repair. Complications occurred in 2 patients (33%) who underwent diversion. Both complications were associated with infection and dehiscence of the midline abdominal incision. The patients without

696 Morken et al

Surgery October 1999

Table V. Management data for group I vs group II Parameters

Group I (n = 13)

Group II (n = 32)

31 23 46 33 67 31 100

65* 59* 91 25 75 25 100

Distal rectal washout (%) Damage repair (%) Diversion of fecal stream (%) Loop colostomy End colostomy Drainage of presacral space (%) Perioperative broad-spectrum antibiotics (%) *P = < .05.

Table VI. Outcome data for group I vs group II Parameters Complications (%) Hospital length of stay (days) Mortality (%) Reversal of colostomy (%) Timing of colostomy reversal (mo)

Group I (n = 13)

Group II (n = 32)

15 7±6 0 86 1.8 ± 1.3

41* 24 ± 19* 6 47* 4.3 ± 1.8*

*P = < .05.

diversion had no complications. The patients who had their fecal stream diverted had a significantly longer hospital length of stay (12 ± 6 vs 3 ± 2 days, P = < .05). Five of the 6 patients (83%) who were diverted in group I had their colostomy reversed at 1.8 ± 1.3 months after injury. DISCUSSION The treatment of civilian rectal injuries has evolved out of experiences gained during wartime. During the Civil War, wounds to the abdomen were treated nonoperatively, with a mortality of 90%.1 World War I brought operative management of colon injuries, with an associated mortality of 58% for selective primary repair.2 Mortality in World War II dropped to 35% after the Army’s Surgeon General made colostomy mandatory for all colonic injuries.3 During the Vietnam conflict, there was a decrease in mortality from 22% to 0% and morbidity from 72% to 10% with the use of presacral drainage and distal rectal washout.4 Today, the reported mortality rates in civilian injuries to the rectum range from 0% to 10% with an associated morbidity of 10% to 45%.7-12 Current treatment of rectal injuries includes various combinations of what we refer to as the four D’s of rectal trauma: (1) distal rectal washout, (2) damage repair, (3) diversion of the fecal stream, and (4) drainage of the presacral space. The use of all these therapeutic maneuvers is accepted in high-velocity wartime injuries. However, the treatment of civilian rectal injuries is a subject of great debate. Diversion of the fecal stream proximal to the site of rectal injury remains the gold standard

for the treatment of civilian rectal trauma.7,15,17 Recently, the first prospective randomized study in the field of civilian rectal trauma was performed to evaluate the role of presacral drainage.15 These authors concluded that placement of a presacral drain had no effect on infectious complications in penetrating rectal injuries. Repair of the damaged rectum is another element in the management of civilian rectal injuries. The consensus in the literature favors repair of the injury if it is accessible, if rectal mobilization is done for exposure of other structures, if large lacerations are not associated with significant rectal wall loss, and, possibly, in patients with associated genitourinary injuries.7,8,18 Distal rectal washout is the most controversial component in the management of civilian rectal injuries. There are many authors who support its routine use,7,10,12 whereas others believe that their results do not support the routine use of distal rectal washout, because no benefit was seen in terms of pelvic sepsis or outcome.8,11,14,18,19 Although there is considerable debate regarding the use of distal rectal washout in civilian rectal injuries, many authors acknowledge its benefits in high-velocity injuries resembling wartime injuries.4,9,11,18,19 Other options besides the dogma of the four D’s have been proposed in the management of civilian rectal trauma. Observation of the injury without surgical intervention, as well as primary repair of rectal injuries without diversion, distal rectal washout, or placement of a presacral drain, has been reported in selected instances.13,14,19 In 1977, Sohn et al20 reported 6 patients with rectal bleeding caused by mucosal laceration without concomitant

Morken et al 697

Surgery Volume 126, Number 4

Table VII. Subgroup analysis of group I: no fecal stream diversion (n = 7) vs fecal stream diversion (n = 6) Parameters

No fecal stream diversion (n = 7)

Age (y) Mechanism of injury (% penetrating) Admission Glasgow Coma Score Systolic blood pressure < 90 mm Hg on admission (%) Injury severity score Number of associated injuries Time to definitive care (min)

31 ± 8 71 14 ± 0.4 16 10 ± 5 1.3 ± 1.5 95 ± 85

Fecal stream diversion (n = 6) 28 ± 11 83 14 ± 0.5 14 19 ± 9 3.5 ± 1.6* 59 ± 41

*P = < .05.

sphincter injury. Two patients required suture of the laceration for hemostasis, whereas 4 patients had only in-hospital observation. They reported no mortality and little morbidity.20 In 1979, Haas and Fox21 reported 7 cases in which no operation was performed for an isolated gunshot wound to the rectum. There were no other organs injured and all wounds healed without complication. In 1996, Levine et al13 reported 6 patients (including 2 gunshot wounds) who underwent primary repair without diversion, noting no morbidity or mortality. Their selection criteria were as follows: RISS grade 2 or less, no major associated injuries, treatment rendered within the first 8 hours after injury, hemodynamic stability, and no need for extensive rectal mobilization. In 1998, McGrath et al17 attempted to classify and treat rectal injuries based on anatomic consideration. They defined extraperitoneal rectal injuries as those occurring in rectum not covered by serosa. Therefore, the anterior and lateral walls of the upper two thirds of the rectum are, by definition, intraperitoneal and treated as distal colon and closed primarily without diversion of the fecal stream, drainage of the presacral space, or distal rectal washout. They suggested that extraperitoneal injuries to the upper two thirds should be explored, repaired without distal rectal washout, and in selected patients, without diversion of the fecal stream. Extraperitoneal injuries in the lower one third were considered adequately treated with exploration and repair alone, without presacral drainage. If not repaired, the authors suggest presacral drainage. The issue of diversion of the fecal stream was not addressed in extraperitoneal injuries of the lower one third of the rectum. There have been advocates for aggressive therapy regardless of the anatomic location or severity of the injury in civilian rectal trauma.7,10,19 More recently, isolated articles have reported a more conservative approach to management of selected rectal injuries with no increase in mortality and minimal morbidity.13,14,19-21 Aggressive surgical management of all rectal injuries has been questioned, and data regard-

ing the use of specific organ injury severity to guide management are lacking. Therefore, we undertook this study to assess whether the RISS developed by the American Association for the Surgery of Trauma can be clinically useful in defining the level of care and possibly impact the outcome of civilian rectal injury.16 Based on a limited number of patients, our study demonstrates that civilian rectal injury is infrequent. When it does occur, it is usually a full-thickness injury (RISS grade 2, 3, or 4) and often associated with penetrating trauma. As the rectal injury progresses from partial thickness to full thickness, the severity of the trauma to the patient increases as does use of hospital resources, mortality and morbidity. In regard to the RISS, our study suggests that more conservative therapy is warranted for RISS grade I rectal (partial thickness) injuries as proposed by Levine et al.13 This approach, in our study, was associated with no mortality and a low incidence of complications (15%). The only complications occurred in those patients receiving diversion of the fecal stream associated with distal rectal washout and placement of presacral drains. Aggressive treatment plans in RISS grade I injuries may be associated with an increased morbidity because they subject patients to what are possibly unnecessary procedures. However, RISS grade 2, 3, and 4 (full thickness) injuries may continue to undergo repair of the injury when possible, distal rectal washout, and diversion of the fecal stream. In our study, this algorithm was associated with an overall mortality of 4% and a morbidity of 33%. In light of the recent prospective data regarding the placement of a presacral drain and our own experience, this practice should probably be used selectively. Like all clinical studies, this study has flaws. This study is limited by the small sample number, it is retrospectively dependent on documentation from the medical record and a trauma registry, it was conducted at a single institution, and the RISS scores have been retrospectively calculated. Last, we lack data on documented colostomy closure. This is due to the transient population involved in

698 Morken et al

penetrating trauma and our policy, as a trauma referral center, of transferring patients back to their community for completion of care. In conclusion, our review contains a small number of patients, and it would be hazardous to draw any firm conclusions from these data. However, our data suggest that extraperitoneal partial-thickness (RISS grade I) civilian rectal injuries can be managed with or without repair and do not require diversion of the fecal stream, distal rectal washout, or placement of a presacral drain. Our patients with full-thickness injuries (RISS grade 2, 3, and 4) were routinely treated with 3 of the 4 D’s of rectal trauma (diversion of the fecal stream, distal rectal washout, and damage repair when possible), making conclusions regarding efficacy impossible. We cannot support the routine placement of presacral drains in civilian rectal injuries, regardless of injury severity. The RISS, as developed by the OISC-AAST, is intended to facilitate clinical research in the field of trauma and is to be refined as clinical experience evolves. Based on our study, the RISS can impact the level of care provided to civilian rectal injury and, if used to guide the management of these injuries, may improve outcome. To assess the utility of these treatments in full-thickness injuries, a prospective randomized multicenter trial would be necessary to provide an adequate population. The management of colonic injuries has clearly moved toward primary closure without diversion, suggesting the possibility of similar treatment for rectal injuries. This study should be based on nomenclature that not only defines the location of the injury, as suggested by McGrath et al,17 but also on the severity of the rectal injury as described by the RISS. If this can be accomplished, perhaps we can develop algorithms based on location and severity of the civilian rectal injury that will provide surgeons with management guidelines and lead to improvement in morbidity without affecting mortality. REFERENCES 1. Elkin DC, Ward WC. Gunshot wounds of the abdomen: a survey of 238 cases. Ann Surg 1943;118:780-70. 2. Wallace C. A study of 1200 cases of gunshot wounds of the abdomen. Br J Surg 1917;4:679-743. 3. Ogilvie WH. Abdominal wounds in the western desert. Surg Gynecol Obstet 1944;78:225-38. 4. Lavenson GS, Cohen A. Management of rectal injuries. Am J Surg 1971;122:226-30. 5. Lung JA, Turk RP, Miller RE, Eiseman B. Wounds of rectum. Ann Surg 1970;172:985-90. 6. Armstrong AG, Schmitt HJ, Patterson LT. Combat wounds of the extraperitoneal. Surgery 1973;74:570-4. 7. Grasberger RC, Hirsch EF. Rectal trauma: a retrospective analysis and guidelines for therapy. Am J Surg 1983;145: 795-9.

Surgery October 1999

8. Burch JM, Feliciano DV, Mattox KL. Colostomy and drainage for civilian rectal injuries: is that all? Ann Surg 1989;209:600-11. 9. Shannon FL, Moore EE, Moore FA, McCroskey BL. Value of distal colon washout in civilian rectal trauma: reducing gut bacterial translocation. J Trauma 1988;28:989-94. 10. Mangiante EC, Graham AD, Fabian TC. Rectal gunshot wounds: management options in penetrating rectal injuries. Am Surg 1986;52:37-40. 11. Tuggle D, Huber PJ. Management of rectal trauma. Am J Surg 1984;148:806-8. 12. Vitale GC, Richardson JD, Flint LM. Successful management of injuries to the extraperitoneal rectum. Am Surg 1983;49:159-62. 13. Levine JH, Longo WE, Pruitt C, Mazuski JE, Shapiro MJ, Durham RM. Management of selected rectal injuries by primary repair. Am J Surg 1996;172:575-9. 14. Thomas DD, Levison MA, Dykstra BJ, Bender JS. Management of rectal injuries: dogma versus practice. Am Surg 1990;56:507-10. 15. Gonzalez RP, Falimirski ME, Holevar MR. The role of presacral drainage in the management of penetrating rectal injuries. J Trauma 1998;45:656-61. 16. Moore EE, Cogbill TH, Malangoni MA, Jurkovich GJ, Champion HR, Gennarelli TA, et al. Organ injury scaling, II: pancreas, duodenum, small bowel, colon, and rectum. J Trauma 1990;30:1427-9. 17. McGrath V, Fabian TC, Croce MA, Minard G, Pritchard FE. Rectal trauma: management based on anatomic distinctions. Am Surg 1998;64:1136-41. 18. Levy RD, Strauss P, Aladgem D, Degiannis E, Boffard KD, Saadia R. Extraperitoneal rectal gunshot injuries. J Trauma 1995;38:273-7. 19. Ivatury RR, Licata J, Gunduz Y, Rao P, Stahl WM. Management options in penetrating rectal injuries. Am Surg 1991;57:50-4. 20. Sohn N, Weinstein MA, Ganchar J. Social injuries of the rectum. Am J Surg 1977;134:611-2. 21. Haas PH, Fox TA Jr. Civilian injuries of the rectum and anus. Dis Colon Rectum 1977;2:17-23.

DISCUSSION Dr Fred A. Luchette (Cincinnati, Ohio). You have done a thorough and in-depth review of your institution’s experience with managing traumatic rectal injuries during the past decade. You conclude that grade I rectal injuries, a “partial thickness laceration,” had greater morbidity when managed with routine fecal diversion. This conclusion is not surprising and intuitively makes sense to all of us. My first question deals with your 2 study groups. The mechanism of injury for the majority of both study groups I and II is classified as penetrating injuries. Do you have any further details about whether these were relatively high kinetic energy wounds such as gunshot injuries or low velocity stab wounds? A difference may explain your findings of improved outcome in the group managed without diversion. Do you know whether your trauma surgeons vary their management of rectal wounds by mechanism? In other words, do your findings merely represent good

Surgery Volume 126, Number 4

clinician judgment due to different penetrating forces? The higher injury severity score and significantly greater number of associated injuries in the group receiving fecal diversion would suggest that there might be some selection bias present. The second question deals with your technique of diagnosing a rectal injury. Digital rectal examination or proctosigmoidoscopy or both determined this. What is the accuracy of these techniques? Is it possible that the individuals in group I receiving fecal diversion who developed infectious complications did indeed have a grade 2 or higher wound and not just a partial-thickness laceration? Thus, the conclusion that fecal diversion is not warranted for grade 1 is not valid because the rectal injury was actually a full-thickness rather than partialthickness wound. Third, are the differences in morbidity between groups I and II purely due to a greater frequency of blunt injury and a higher incidence of gunshot wounds in comparison with group I? The significantly higher injury severity score and number of associated injuries would suggest this difference in mechanism might have influenced the decision for fecal diversion or no diversion. Or is the increased morbidity due to the infrequent use, only 25%, of presacral drainage in group II? My last question concerns the low frequency, only 47%, of colostomy closure in group II. My experience is that these patients are usually anxious to get rid of colostomy. Why did only 47% of the group II patients, that is, those with a grade 2, 3, or 4 injury, undergo closure of the colostomy? Were the injuries not healed or were there other patient issues? I agree with your findings; however, I believe that most of your observations may reflect good clinical judgment of the trauma surgeons at Hennipen County. Dr Morken. The incidence of gunshot wounds was equal between groups. The full-thickness injury group did have a higher incidence of blunt trauma, 31% compared with 8% in our partial-thickness group. Do our surgeons treat these injuries based on mechanism? The answer is no. However, they do treat the injuries based on the injury severity score and number of associated injuries, which likely relate back to the mechanism of injury. Regarding your second question on the accuracy of diagnosing a rectal injury and then computing the RISS grade, I think the honest answer is that no one knows. The literature reports sensitivity, but no one discusses accuracy. Sensitivity is anywhere from 75% to 95% for digital rectal examination alone and proctoscopy alone is reported to be anywhere from 47% to 100% sensitive. Our values fell within these ranges. Your third question addressed the difference in morbidity between our 2 groups. We think it is likely due to the higher incidence of blunt trauma resulting in more

Morken et al 699

associated injuries than our partial-thickness group. Regarding the use of presacral drainage specifically, there was no difference in its use between our 2 groups; they both had it infrequently. Your last question addressed the low incidence of colostomy reversal in our study. I think this merely reflects our patient population. Our data represented documented reversal occurring in our facility. As a large trauma center, we get many patients in referral who then return to the outside facility, so we don’t know whether they had reversal or not. We also have a somewhat transient patient population. Dr Mark A. Malangoni (Cleveland, Ohio). You are right to point out the problem with injury location and the scale to evaluate these rectal injuries, and that is a very important criticism of the scale. Can you tell us your opinion about the role of resection and colostomy in patients who have high rectal injuries? These are extraperitoneal injuries that you can reach from within the abdomen by mobilizing that upper portion of the rectum. Do you think there is a role for resection in those injuries? In our experience, they are very difficult to repair. Dr Morken. In our study, full-thickness injuries in that area underwent colostomy. If they had massive damage in that area not amenable to repair, they underwent resection of that portion of the rectum with a Hartmann’s procedure. Dr T. J. Howard (Indianapolis, Ind). My question is on the distance from the anorectal verge for these injuries. Is that variable included in the severity score? Colorectal surgeons and surgical oncologists often do full-thickness rectal excisions without a diverting colostomy, and those patients would fall into your group II patients who get a mandatory colostomy in trauma. Could you comment on these thoughts? Dr Morken. The RISS as developed by this committee does not take into account the injury’s distance from the anal verge. It mainly looks at the thickness and size of the injury. Dr Frank R. Lewis (Detroit, Mich). I still don’t understand your classification of the group I injuries. By definition they are partial thickness. Also by definition of your subgroup, the injury had to occur from the outside, because they are virtually all gunshot wounds and some presumably pelvic fractures. So these are not injuries occurring from the luminal side. Thus by definition if it is partial thickness starting from the outside, you can only see it from the outside. That would require intraperitoneal exploration of the rectum, which was not done in all of these people. I don’t understand exactly what the criteria were for knowing that there is a partial-thickness injury present. Could you explain that just a bit more? In patients who were not explored and in those who were explored. Did

700 Morken et al

you fully mobilize down into the pelvis to see these? I wasn’t clear from your presentation exactly how that went. So again it gets to Dr. Luchette’s question about accuracy of classification. Dr Morken. In answer to the first part of your question, our study did have a significant number of patients who sustained significant trauma to the rectum via an assault. Several patients had broomsticks forcibly placed in an assault. Therefore, some injuries were intrarectal in nature. Also, some gunshot wounds were partial thickness in nature because they inflicted glancing blows or blast injury to the rectum.

Surgery October 1999

Did we fully mobilize the rectum in all instances? The answer is no, because we do not have a protocol for these injuries instituted at our facility at this time. The question regarding accuracy of diagnosis is a valid one. We don’t necessarily have the answer to that. With proctoscopy and intraoperative examination we thought that we were able to tell the difference between a full- and a partial-thickness injury based on the surgical reports that were filed.