Approximately 20% of patients initially undergoing nonoperative management of blunt splenic injury require further intervention. Failure has been associated with the presence of a contrast blush in up to two-thirds of these patients.68 The presence of a contained contrast blush within the parenchyma of the spleen represents pseudoaneurysm formation of a branch of the splenic artery. Angioembolization is now commonly used to selectively occlude the arterial branches containing these injuries.61,62,65,69,70 Implementation of this salvage technique at centers that routinely screen for the presence of pseudoaneurysm has increased the success of nonoperative management to 90% or greater. Pseudoaneurysm formation has been observed in even grades I and II injuries and may not be present on the initial imaging.61,64,70 Therefore, follow-up CT scan is recommended on all patients with splenic trauma within 24–48 hours after injury. If these images show stable injuries without pseudoaneurysm formation, expectant management may ensue. Figure 12-7 is a management algorithm for the patient with blunt splenic injury.
Long-term data are unavailable concerning the risk of outpatient or delayed rupture, but the incidence is low and has been reported to be about 1.4%.71 The average date to readmission for delayed splenectomy after discharge was 8 days in this study. Lower-grade (I, II) injuries tend to heal more quickly and most all injuries are healed by 5–6 weeks.72 However, approximately 20% of blunt splenic injuries will not show complete healing and may be at risk for pseudocyst formation. A CT scan should be repeated in 6 weeks for grades I and II injuries and 10–12 weeks in grades III–V before reinstating normal activity.
Patients requiring urgent or emergent intervention for splenic hemorrhage may develop hypothermia, coagulopathy, and visceral edema. The most expeditious and safest course of action under these conditions is removal of the spleen. The general assumption of abdominal exploration for trauma is that there are known and, possibly, unknown injuries. The operative approach is via a midline vertical incision that allows the best exposure and facilitates temporary abdominal closure should visceral edema or damage-control measures be necessary. Standard operating procedure is similar to that previously highlighted in the section, Management of Penetrating Abdominal Trauma.
With respect to performing a splenectomy, a Buckwalter retractor is used to expose the left upper quadrant. The spleen is retracted medially with some downward compression, and the posterior attachments can be taken with the cautery. Once these attachments are freed, the spleen can be mobilized medially for optimal exposure. The assistant stands on the left side of the table and supports the spleen while the surgeon ligates short gastric and hilar vessels. Being careful to avoid the tail of the pancreas, a large clip, placed on the specimen side of the splenic hilum, will reduce back-bleeding and expedite the procedure. Once the spleen has been removed, the splenic fossa is inspected for further bleeding with a rolled laparotomy pad.
Hemodynamically stable patients found to have small to moderate amounts of parenchymal hemorrhage at laparotomy may be candidates for splenic preservation. The spleen is mobilized into the wound using the same technique as for splenectomy. The injury to the spleen is assessed and decision is made whether to resect a portion if the parenchymal injury extends into the hilum or if arterial bleeding is coming from within the splenic laceration itself. If the decision is made to resect the upper or lower pole, the parenchyma is divided with the cautery, and the associated hilar vessels are taken with clamps and ties. Any arterial bleeding from the parenchyma is controlled with suture ligature and the cautery is used to control oozing from the parenchyma. A tongue of omentum is then sutured into the laceration or to the raw surface of the remaining spleen in the case of resection. Approximately 50% of the spleen is required to preserve adequate phagocytic and immunologic function. If this cannot be achieved, a splenectomy is probably the best option.
Overwhelming Postsplenectomy Infection.
The incidence of overwhelming postsplenectomy infection (OPSI) following trauma is not well understood because it may not be appreciated when it occurs and it is not routinely reported. However, the reported incidence of OPSI in adult patients undergoing splenectomy for all causes is 0.9% with a mortality of 0.8%.73 The risk of OPSI in adults following trauma is felt to be lower than the incidence seen after splenectomy for hematological disorders such as idiopathic thrombocytopenic purpura (ITP), lymphoma, and thalassemia. Children are at greater risk for OPSI and should receive prophylactic penicillin V 125 mg twice daily until age 3 and then 250 mg twice daily until age 5. Currently, anyone older than 2 years should receive the 23 valent pneumococcal vaccine and a one-time dose of the Haemophilus influenzae and meningococcal vaccine. A one-time booster dose of the pneumococcal vaccine is recommended 5 years after the original vaccine.74
The liver is susceptible to the same blunt force mechanisms as the spleen making it the second most frequently injured intra-abdominal organ. Similar to the spleen, nonoperative management of blunt liver injury has greatly reduced transfusion requirements, hospital length of stay, and mortality.75–78 Angiographic embolization of arterial injury has also greatly reduced the morbidity and mortality of liver trauma. Complications of nonoperative management such as biloma and liver abscess can usually be managed with minimally invasive techniques as well. What is imperative in the management scheme is the knowledge of when to take the patient immediately to the operating room for active hemorrhage versus attempting nonoperative management with angioembolization. The liver is obviously not as expendable an organ as the spleen, and there is no substitute for an in-depth knowledge and experience handling hepatic injuries.
Similar to the experience with blunt splenic trauma, routine use of CT scans has revolutionized the management of blunt hepatic trauma. The most recent data show that 70–85% of all patients presenting with blunt liver trauma can be managed nonoperatively.75,76,79,80 Patients must be hemodynamically stable, free of peritoneal signs or other evidence of bowel injury, and have absence of contrast extravasation. Worsening grade of injury makes nonoperative management less likely, but injury grade alone should not dictate the decision to intervene.78,81 Patients should have a stable systolic blood pressure greater than 90 mm Hg with a heart rate less than 100 beats/min after controlling other possible sources of extra-abdominal blood loss, such as orthopedic and soft tissue injuries. Failure from subsequent liver hemorrhage occurs in 0.4–5% of patients and failure due to missed injury of other intra-abdominal organs, such as the kidney, spleen, pancreas, and bowel, occurs in 0.5–15% of patients.75,76,79,81,82 These data are summarized in Table 12-8. It is difficult to tell the specific cause of immediate surgery in the earlier reports because laparotomies for all causes, such as associated splenic hemorrhage, were included. Christmas et al and Velmahos et al report that rates of immediate operative intervention for liver hemorrhage are 15 and 13%, respectively.78,79
Table 12-8: Failure Rates for Nonoperative Management of Liver Trauma ||Download (.pdf)
Table 12-8: Failure Rates for Nonoperative Management of Liver Trauma
|Study (year)||Number of Patients||Immediate Surgery (%)||Overall Failure Rate (%)||Liver Failure Rate (%)||Other Failure Rate (%)|
|Meredith et al81 (1994)||116||48||3||3||0|
|Pacther et al82 (1996)||404||53||1.2||0.7||0.5|
|Velmahos et al79 (2003)||78||29a||15||0||15|
|Christmas et al78 (2005)||561||32b||1.8||0.4||1.4|
Angiographic embolization is a useful adjunct in the management of blunt hepatic trauma both in nonoperative patients and those who have undergone damage-control laparotomy in a number of small series.78,79,83–89 Intravenous contrast extravasating from the liver parenchyma into the peritoneal cavity and contrast contained within the parenchyma of the liver associated with a large amount of intraperitoneal blood on initial CT scan are emergent situations mandating angiographic embolization or surgery.90 If bleeding appears to originate from the retrohepatic vena cava or hepatic veins and is ongoing, emergent exploration is the only choice because arterial embolization is ineffective in these scenarios. Patients with labile blood pressure and parenchymal injury may be better served with emergent exploration depending on the time necessary to assemble the angiography team. However, if angiography is readily available, favorable results have been obtained transporting patients to the angiography suite with ongoing resuscitation for arterial bleeding.86,91 Patients requiring significant resuscitation during a successful embolization procedure may still be at risk for abdominal compartment syndrome and should have bladder pressures monitored if they receive greater than 10 units of blood products.79,92
Patients with contrast extravasation contained within the liver parenchyma without significant associated hemoperitoneum are less worrisome but should probably undergo arteriogram for further assessment.90 A perceived contrast blush on CT scan is associated with bleeding identified at angiography in approximately 60% of patients.87 Arteriography in the absence of contrast extravasation will likely be a low-yield study. Patients requiring operative exploration of their liver wounds that have arterial bleeding from deep within the liver parenchyma may benefit from hepatic packing with radiolucent laparotomy pads and direct transportation to the angiography suite.83,84,88
Several complications can occur in the management of liver trauma, including biloma, hepatic necrosis, liver abscess, and gallbladder necrosis.85 Minimally invasive techniques can be successfully used to handle the majority of these complications. Biloma can occur after significant damage to the biliary tree. Patients may present with an ileus, abdominal pain, distention, or an abscess often with some degree of hyperbilirubinemia. Percutaneous drainage for localized collections will often control the leak and immediate symptomatology.87,88 Persistent bile leak will usually require an endoscopic retrograde cholangiopancreatography (ERCP) with biliary stent placement. Patients with hepatic duct or common bile duct injury may require hepatectomy or hepaticojejunostomy. Patients with generalized ascites may benefit from laparoscopic washout and placement of perihepatic drains followed by ERCP for persistent leak.78,88
Simple abscess can usually be managed with percutaneous drainage, unless significant hepatic necrosis simultaneously exists.88 Sterile liver necrosis may eventually resolve with expectant management but usually requires hepatic debridement when associated with infection. Patients who undergo embolization without surgery have approximately a 20% chance of developing infected necrosis, while patients who undergo laparotomy followed by hepatic artery embolization have an incidence of hepatic necrosis of over 80% in one series.85 Gallbladder necrosis occurs in approximately 20% of grades IV and V injury following nonselective embolization of the right hepatic artery.87 These patients usually present several days after injury with leukocytosis, hyperbilirubinemia, and abdominal pain. HIDA (hepatobiliary iminodiacetic acid) scan and a high index of suspicion may be necessary to differentiate this diagnosis from the others listed.
Bleeding from minor liver injuries (grades I and II) usually stops spontaneously, and surgical intervention is rarely required.75,81,93 Occasionally, patients may require exploration for other injuries after abdominal trauma in the presence of minor liver trauma. Nonbleeding liver injuries should be left alone. In the face of coagulopathy or hypothermia, minor hepatic injuries may present with persistent oozing. In such cases, topical hemostatic agents, with or without perihepatic packing, may be all that is necessary to stop the bleeding if the patient is being adequately resuscitated.
Major liver injuries (grades III–V) are more likely to bleed and require surgical intervention. Because grades IV and V liver injuries can present formidable technical challenges even in the hands of the most capable individuals, a variety of surgical techniques have been developed for their management.
Large liver wounds should be quickly inspected to get some idea about the degree of hemorrhage and then packed off initially. Anesthesia should be notified about anticipated blood loss and blood availability checked. Vital signs and resuscitation status should be reviewed. Bleeding should be contained with packing and direct pressure until anesthesia has had time to “catch up.” Remaining focused and well organized, as well as replenishing intravascular volume, is invaluable in the management of any trauma patient.
The most direct approach at this point is to remove the packing and visually inspect for bleeding vessels which can be individually ligated. Debridement of devitalized tissue using finger fracture technique will expose additional bleeding vessels which may have retracted into the surrounding parenchyma. If bleeding prevents adequate visualization of the surgical field, the next step should be vascular control of the portal triad (ie, the Pringle maneuver).93 This maneuver is easiest to perform for the person standing on the left of the patient. This individual places the fingers of the left hand into the foramen of Winslow and uses the thumb to palpate for the cord of tissue running into the caudal surface of the liver. Once this structure is suspected, its identity can generally be confirmed by appreciating the pulse in the hepatic artery. A hole is then created in the hepatoduodenal ligament using blunt finger dissection. A noncrushing vascular clamp can be applied or a ½-in. Penrose drain can be doubled, looped around the porta and cinched down with a Kelly clamp. We prefer the latter technique because it seems to be less obtrusive to further manipulation of the liver and may be less traumatic to the structures of the porta.
If a Pringle maneuver does not adequately decrease liver bleeding, concern for hepatic vein or retrohepatic caval injury should be entertained. Obtaining adequate exposure in deep liver wounds or in juxtahepatic caval injuries is of utmost importance. The falciform ligament is taken off the diaphragm posteriorly to the bare area. The right and left triangular ligaments are dissected with the cautery extension to the corresponding coronary ligaments. Further dissection of the coronary ligaments to the bare area will allow vigorous mobilization of the liver into the surgical field. Careful dissection of the bare area will allow access to the suprahepatic inferior vena cava. If the plane in the bare area is difficult to develop, a transverse incision in the diaphragm here will gain access to the pericardium and intrapericardial control of the inferior vena cava can be achieved.94 Total hepatic isolation can be achieved with a Pringle maneuver and vascular control of the infrahepatic, suprarenal inferior vena cava, and the suprahepatic intra-abdominal or intrapericardial inferior vena cava. The final step in this procedure is occluding inflow to the intra-abdominal aorta at the diaphragm while cardiac return from the lower body is eliminated. The physiologic stress of this technique may not be well tolerated in patients with severe shock and clamp times less than 20–30 minutes should be maintained.
Vascular shunts of the liver, otherwise known as atriocaval shunts, have been devised to circumvent impeding cardiac return when vascular isolation of the liver is desired.95 In this procedure, early recognition of juxtahepatic venous injury is essential and liver bleeding is temporarily contained with packing and direct pressure.96 A two-team approach is best; a median sternotomy is performed simultaneously while gaining vascular control of the infrahepatic suprarenal vena cava with a Rumel tourniquet. A Rumel tourniquet is also necessary around the intrapericardial inferior vena cava. A purse-string suture is placed in the right atrial appendage, and an opening is created in this structure. A 32F chest tube can then be directed from the atrium into the inferior vena cava. The Rumel tourniquets are applied. Before inserting the chest tube, an additional fenestration should be created approximately 3 in from the tapered proximal end. This opening will allow egress of the blood from the chest tube returning from the lower half of the body back into the heart. The proximal end of the chest tube protrudes from the right atrial appendage that is secured with the purse-string suture. The end of the chest tube is clamped or can be accessed with a Christmas tree adapter and used as a large-bore resuscitation line. Exploration of the retrohepatic cava is now possible. A number of institution-specific variations of this technique have been described, but we are unaware of any technique that has demonstrated improved results over Schrock's original description.97–99
The alternative to shunting grade V injuries is termed “direct exposure” popularized by Pachter and others.100–102 This technique consists of four basic principles: (1) manual compression and packing of the liver with vigorous resuscitation, (2) portal triad occlusion, (3) wide mobilization of the hepatic ligaments allowing for medial rotation of the liver and exposure of the retro hepatic cava, and (4) extensive finger fracture including normal liver parenchyma for direct vascular control of the injury.102 In one analysis of over 142 patients sustaining juxtahepatic venous injuries, 35 (24.6%) were managed without a shunt, resulting in a survival rate of 49%. The patients managed with a shunt had a survival rate of 19%.102 Suffice it to stay, unless there is a significant institutional experience with the shunt, direct exposure or total hepatic isolation alone may provide the best chance for patient survival for the occasional patient presenting with a grade V injury.
Lobar hepatic resection offers exposure of the involved hepatic veins and retrohepatic vena cava in grade V injuries but has largely fallen out of favor because of high associated mortality rates.96,103–106 The only current recommendation for this procedure is when the majority of the lobectomy occurred at the time of injury and the disrupted portion or lobe of liver has questionable viability. Selective hepatic artery ligation is another technique that has been employed to control arterial bleeding deep within the liver that cannot be easily identified or controlled through the liver wound.107–110 In this situation, if a Pringle maneuver greatly reduces the arterial bleeding, the artery to the respective lobe should be dissected out and occluded. If this maneuver maintains hemostasis, the vessel may be taken. A more recent option is to pack the patient with radiolucent laparotomy pads and take him/her for an arteriogram and possible embolization.83,84,88
Once patients with operative liver trauma sustain surgical hemostasis, they may become hypothermic and coagulopathic with bleeding occurring from nonsurgical sources, in particular the raw liver parenchyma. At this point, the liver and other sources of nonsurgical bleeding may be packed with laparotomy pads and a temporary abdominal closure performed.106,111–116 Patients can then be transported to the intensive care unit where they may be further resuscitated and warmed. Take back for removal of the packing, and debridement of devitalized liver may generally be undertaken safely in 24–48 hours. Omental packing of the liver defect originally described by Stone may reduce the incidence of bile leak and abscess formation.117
There is yet no place for nonoperative management of hollow viscus injury, and the nemesis of nonoperative management of blunt abdominal trauma is therefore the missed bowel injury and all its catastrophic consequences. Otherwise, most management is straightforward—debridement and primary repair for nondestructive injuries and resection with primary repair versus stomal formation for destructive injuries.
Radiographic Findings of Blunt Bowel Injury
There are two basic types of findings of bowel injury on CT scan: direct and indirect. Direct findings are usually straightforward if present and amount to extravasation of oral contrast (if administered) and free air, which have been reported to occur in 4 and 28% of the time, respectively. Little else can explain the first of these two entities, while free air from other sources such as extensive subcutaneous emphysema tracking through a diaphragmatic hiatus is unusual.118–120 Indirect findings may be subtle and can vary in presentation depending on the quality of the scan. Indirect findings include mesenteric hematoma or contrast blush, bowel wall edema, unexplained free fluid, “fat streaking” and bowel loops that don't opacify with intravenous contrast (Table 12-9).
Table 12-9: CT Scan Findings of Blunt Bowel Injury ||Download (.pdf)
Table 12-9: CT Scan Findings of Blunt Bowel Injury
|Oral contrast extravasation||Mesenteric hematoma|
|Free air||Mesenteric blush|
|Bowel wall edema|
|Unopacified (vascular contrast media) bowel loops|
Mesenteric hematoma is nonspecific and can occur from associated injuries, such as pelvic fractures or renal injuries with hematomas from these structures tracking into the bowel mesentery. However, a vascular blush in the leaves of the mesentery is indicative of active hemorrhage until proven otherwise and, generally, a determinant for immediate operative exploration. Bowel wall edema and ascites are common in blunt trauma patients, can occur from resuscitation of other injuries, and don't necessarily connote bowel injury. Free fluid in the absence of solid organ injury can be further evaluated with DPL if the abdominal examination is unreliable. Fat streaking can occur with mesenteric contusion and does not necessarily portend an operative indication. Unopacified bowel loops can indicate vascular disruption of the mesentery or simply be due to poor contrast timing in an under-resuscitated patient. In review of 8112 CT scans, Malhotra showed that at least one of these findings was present in 88.3% of patients with blunt bowel or mesenteric injury and that the likelihood of finding an injury at exploration increased when there was an increasing number of these findings.
Appreciation of the AAST organ injury grading scale is helpful in describing wounds of the bowel.121 Grade I injuries are contusions and partial-thickness lacerations of the bowel wall without perforation. Grade II injuries are full-thickness wounds involving less than 50% of the bowel wall circumference. Grade III are lacerations comprising greater than 50% of the bowel wall circumference without complete transaction. Grades IV and V injuries represent complete transection of the bowel wall and transection with segmental tissue loss and/or devascularization of the mesentery respectively. The terms destructive and nondestructive simplify the terminology; nondestructive wounds are those injuries that can be managed with debridement and primary suture enterorrhaphy and comprise grades I–III.122 Destructive wounds require resection of an entire segment of the bowel due to loss of colonic integrity or devascularization of the mesentery and encompass grades IV and V (Tables 12-10 and 12-11).
Table 12-10: AAST Small Bowel Injury Scale ||Download (.pdf)
Table 12-10: AAST Small Bowel Injury Scale
|Gradea||Type of Injury||Description of Injury||ICD-9b||AIS-90c|
|I||Hematoma||Contusion or hematoma without devascularization|
|Laceration||Partial thickness, no perforation||2|
|II||Laceration||Laceration <50% of circumference|
|III||Laceration||Laceration ≥50% of circumference without transection||3|
|IV||Laceration||Transection of the small bowel|
|V||Laceration||Transection of the small bowel with segmental tissue loss|
Table 12-11: AAST Colon Injury Scale ||Download (.pdf)
Table 12-11: AAST Colon Injury Scale
|Gradea||Type of Injury||Description of Injury||ICD-9b||AIS-90c|
|I||Hematoma||Contusion or hematoma without devascularization|
|Laceration||Partial thickness, no perforation|
|II||Laceration||Laceration ≤50% of circumference|
|III||Laceration||Laceration >50% of circumference|
|IV||Laceration||Transection of the colon|
|V||Laceration||Transection of the colon with segmental tissue loss|
The distinction between destructive and nondestructive wounds is important in terms of the prescribed management. Nondestructive wounds of the large or small bowel can generally be repaired without further consideration. Most small bowel destructive injuries should be resected and reconstituted unless damage-control conditions prevail.
In contrast to the small bowel, the management of colon injuries has received great scrutiny. Ushering in the dawn of modern-day trauma surgery, the WWII military experience dictated that all colon wounds, destructive or not, be managed by colostomy. This philosophy remained surgical dogma until the 1980s.123,124 In a comprehensive review of the literature since 1979, primary repair of the colon for nondestructive wounds had been shown to have a leak rate of 1.6%.122 Compared to patients receiving colostomy for similar types of wounds, the incidence of intra-abdominal abscess was 4.9% for primary repair and 12% for colostomy, and overall complication rate was 14% for primary repair and 30% for colostomy. Mortality rates were similar at 0.11% for primary repair and 0.14% for colostomy. These findings clearly show the superiority of primary repair for nondestructive wounds of the colon.
Several risk factors for anastomotic failure pertaining to destructive colon injury have been addressed in the literature: hypotension, shock, interval from injury to operation, amount of fecal contamination, associated organ injury, transfusion requirement, and comorbid disease.125 No data have conclusively shown that any of these risk factors increase the likelihood of anastomotic failure with certain caveats. Patients with massive blood loss or shock may be better served by undergoing a damage-control procedure, with definitive repair delayed.126 Interval from injury to repair greater than 12 hours can be a relative contraindication, if there is wide-spread (greater than one quadrant) fecal contamination. Greater than one- or two-organ system injury has been a concern, but this may just be a marker for degree of shock and overall physiologic derangement. Comorbidities such as AIDS and cirrhosis deserve special consideration, and these patients may be better off with diversion.127,128 Aside from suture line failure, patients with any of these risk factors have a higher incidence of intra-abdominal abscess and overall complications rates.122
Not withstanding the risk factors for colon trauma, most series regard resection and primary anastomosis for destructive colon wounds in a favorable light. In a collective review of 207 patients reported in the literature, management of destructive bowel injury with resection and primary anastomosis had a reported leak rate of 7.2% with a mortality of 1.7% attributable to the colon wound.122 In the largest single institution experience, Murray showed a leak rate of 11% in 112 patients undergoing resection and primary anastomosis for destructive colon wounds with two deaths related to leaks.129
In a multi-institutional trial, Demetriades reported 297 patients with destructive colon wounds in which 197 underwent resection and anastomosis and 100 underwent diversion.127 The choice of operation was left to the discretion of the attending surgeon at the time of exploration. Not surprisingly, the patients with diversion were significantly more injured and ill than those being reconstituted. The anastomotic leak rate was 6.6%, with one leak from the stump of a Hartmann's pouch in the diverted group and four deaths related to anastomotic failure. Multivariate analysis showed no significant difference in mortality or abdominal complications between diversion and primary anastomosis groups. The authors concluded that “patients can be managed by primary repair regardless of risk factors.” This study certainly demonstrates a liberal use of resection and primary anastomosis in a relatively sick and injured cohort of patients. However, the ultimate decision for the choice of operation was up to the discretion of the surgeon at the time of operation, for which there is no substitute.
At laparotomy, the bowel should be examined in its entirety after all other sources of major bleeding are controlled. Small injuries should be noted and tagged with an identifiable suture for easy reference. Larger wounds contributing to ongoing soiling can be temporarily controlled with a “whip stitch” (quick running suture) or Babcock clamps. Mesenteric injuries are identified and active bleeding controlled appropriately. Attention should be directed to the location of the superior mesenteric artery for injuries encroaching on the root of the mesentery. Mesenteric hematomas should be explored with ligation of injured vessels and mesenteric defects closed by careful reapproximation of the peritoneal edges so as not to compromise any feeding vessels. Bowel viability should be noted in relation to any mesenteric injury. Clusters of grade I through III injuries may be resected or individually repaired depending on the particular injury pattern. In blunt trauma, there is usually only one or two grade II or III wounds that can be repaired primarily or one or more devitalized segments that require resection.
Small, superficial grade I injuries can be left alone while deeper, longer grade I injuries can be closed with a simple running suture or interrupted Lembert sutures. Grades II and III wounds should be debrided back to healthy, viable bowel and closed transversely preventing narrowing of the lumen of the bowel. Single layer running or interrupted closure is generally sufficient for repair of the small bowel. When there is significant bowel wall edema, peritonitis or soiling, a two-layer closure with a running inner layer and interrupted Lembert outer layer may be preferential. Grades I and II colon wounds may be managed with single layer closure, but we usually close grade III colon wounds in two layers for added protection.
The leak rate associated with stapled versus hand-sewn anastomosis for destructive wounds of the bowel has been an area of controversy. In two retrospective studies totaling 284 patients undergoing stapled versus hand-sewn anastomosis, Brundage et al showed that hand-sewn procedures had lower leak rates.130,131 Two other retrospective studies totaling 484 patients showed no difference in the leak rate of stapled versus hand-sewn procedures.132,133 Brundage's two studies included 78 colon wounds while the other studies were confined only to the small bowel. Stapled procedures may be a little quicker, particularly if there is more than one anastomosis. In general, the technique chosen according to the literature can be a matter of personal preference. With edematous bowel, the hand-sewn technique is a more prudent approach.