Stomach and Small Bowel
INTRODUCTION
Injuries to the stomach and small bowel are common in penetrating abdominal trauma.
The incidence of gastrointestinal injury following gunshot wounds that penetrate the peritoneal cavity is over 80%. Thus, exploratory laparotomy is warranted on virtually all gunshot wounds that penetrate the peritoneal cavity. The incidence of hollow viscus injury (HVI) secondary to stab wounds that have penetrated the peritoneal cavity is much less, which in most series is about 30%. Thus, a selective approach to operative exploration has been advocated following stab wounds.
Blunt injuries to the stomach and small bowel are much less common than penetrating injury, but collectively compromise the third most common type of blunt abdominal injury. The increasing use of computed tomography (CT) for diagnostic evaluation of the patient with blunt abdominal trauma and selective nonoperative management of solid organ injuries have contributed to some of the difficulties and controversies in the management of HVIs following blunt trauma. In contradistinction to some of the diagnostic difficulties with stomach and small bowel injuries, operative repair of stomach and small bowel injuries is relatively straightforward. The key to the successful management of stomach and small bowel injuries is prompt recognition and treatment, thus decreasing the likelihood of abdominal septic complications and subsequent late death.
HISTORICAL PERSPECTIVE
Intestinal injuries were reported early in the medical literature (see Chapter 1). Small bowel perforation from blunt trauma was first recognized by Aristotle.1 Hippocrates was the first to report intestinal perforation from penetrating abdominal trauma. In 1275, Guillaume de Salicet described the successful suture repair of a tangential intestinal wound. Reports of attempted surgical repair of gastric and intestinal wounds appeared in the literature with heightened interest and controversy during the American Civil War, the Spanish-American War, the Russo-Japanese War, and other military conflicts. However, the dismal results of surgical intervention lead to abandonment of laparotomy even with obvious intestinal injury during these military campaigns.2
By the late 19th century, improved surgical techniques led to renewed interest in laparotomy and repair of penetrating abdominal injuries. Theodore Kocher was the first surgeon to report successful repair of a gunshot wound of the stomach. Although still controversial, in 1901 President William McKinley, shot in the abdomen by an assassin, underwent expeditious transport and surgical repair of several gastric wounds. However, a wound to the pancreas was overlooked, and McKinley died 8 days later.
A laparotomy for intestinal perforation at the start of World War I carried a mortality rate of 75–80%, almost equal to the mortality rate of nonoperative management. However, in the later part of World War I, operative management was recognized as the preferred management for penetrating abdominal trauma.
In World War II, prompt evacuation, improvements in anesthesia, and better understanding and treatment of shock led to mortality rates of 13.9% for jejunal or ileal injuries and 36.3% if multiple injuries were present.3 Further improvements in mortality were noted during the Korean War and Vietnam conflicts. Lessons learned by military surgeons were quickly adopted by surgeons in civilian centers. However, the sheer volume of trauma in several major trauma centers dictated that patients with abdominal stab wounds be treated expectantly and operations performed on the basis of physical signs of peritonitis or hemodynamic instability. Based on this experience, it became obvious that stab wounds and/or low velocity gunshot wounds did not share the same risk as wounds from military weapons. Thus, selective management became a formal policy at larger experienced centers and soon included GSW that appeared tangential, or involved posterior, flank, or thoracoabdominal locations.
Patients with gastrointestinal injuries under these circumstances may benefit from a staged operative approach for abdominal injuries. Simple repairs or temporizing measures aimed at controlling ongoing peritoneal contamination by gastrointestinal injuries are performed at the initial operation. Definitive repair for more complex gastrointestinal injuries is undertaken at a later operation after hemodynamic stabilization.
Nonetheless, prompt recognition and repair of injuries are responsible for the low morbidity and mortality from isolated gastric and small bowel injuries. Associated injuries contribute significantly to morbidity and mortality in patients with gastric and small bowel injuries.
ANATOMY AND PHYSIOLOGY
The stomach generally occupies the left upper quadrant of the abdomen. The nondistended stomach, especially in a supine individual, is located largely in the intrathoracic abdomen where it is offered some protection by the lower chest wall. The position of the stomach can be quite variable and in the erect individual may extend into the lower abdomen, particularly when distended with food or liquid. The stomach is fixed on its lesser curvature by the gastrohepatic ligament, cephalad by the gastrophrenic ligament, and distally by the retroperitoneal duodenum. The greater curvature of the stomach is loosely bound to the transverse colon via the greater omentum and to the spleen by the gastrosplenic ligament. The stomach enjoys a rich blood supply from the left and right gastric arteries, the left and right gastroepiploic arteries, and the short gastric arteries (Fig. 31-1). Venous drainage follows the arterial supply to the stomach for the most part.
FIGURE 31-1 Blood supply to the stomach. An anomalous left hepatic artery can arise as a branch of the left gastric artery. This should be looked for when doing gastric resections. (Reproduced with permission from Agur AMR, Dalley AF, eds. Grant’s Atlas of Anatomy. 11th ed. Copyright Lippincott Williams & Wilkins; 2004.)
The normal stomach is relatively free of bacteria and other microorganisms because of the low intraluminal pH.4 However, up to 103 organisms/mL, including lactobacilli, aerobic streptococci, and even Candida, may be isolated. Low gastric acidity due to H2-receptor blockade or now proton pump inhibitors leads to increased bacterial concentrations in the stomach and proximal gastrointestinal tract, increasing the risk of peritoneal contamination with gastric perforation. Retained food in the stomach may also increase the risk of infection following gastric perforation. Factors include acid neutralization and increased luminal bacteria in the postabsorptive state as well as serving as an adjuvant in the event of peritoneal contamination.
The small bowel distal to the ligament of Trietz is approximately 5–6 m in length in the adult. Protected anteriorly only by the abdominal wall musculature and occupying most of the true abdominal cavity, the small intestine is anatomically vulnerable to injury.
The small intestine is suspended from the posterior abdominal wall by its mesentery, the base of which extends from the duodenal jejunal flexure, superior to inferior and left to right to the level of the right sacroiliac joint. The arterial supply to the small bowel is provided by the superior mesentery artery (SMA), which emerges from under the pancreas and then courses anterior to the uncinate process of the pancreas to enter the root of the mesentery. The blood supply to the small bowel comes from the left side of the SMA via intestinal arteries (Fig. 31-2). The jejunal and ileal branches vary in number and supply all but the terminal part of the ileum. This is supplied by branches from the ileocolic artery. Numerous intestinal arcades form within the mesentery to assure excellent collateral blood supply to the small intestine. Venous return from the small intestine follows the arterial supply: the superior mesenteric vein joins the inferior mesenteric vein and splenic vein to form the portal vein.
FIGURE 31-2 Blood supply to the small bowel. Multiple branches to the jejunum and ileum originate directly from the superior mesenteric artery. The distal ileum is supplied via the ileocolic artery. (Reproduced with permission from Agur AMR, Dalley AF, eds. Grant’s Atlas of Anatomy. 11th ed. Copyright Lippincott Williams & Wilkins; 2004.)
Although no clear distinction exists, the first 40% or so of the bowel is jejunum and the remainder is the ileum (Fig. 31-3). The jejunum has a larger diameter, more circular folds, and larger villi, but less lymphoid tissue than the ileum. The mesentery of the jejunum contains only a single arcade, whereas more than two or three sets of vascular arcades are present in the ileum. Mesenteric fat is also more prominent in the ileum than in the jejunum. The proximal jejunum is the primary site of carbohydrate, protein, and water-soluble vitamin absorption. Fat absorption occurs over a larger length of small bowel. The ileum is the primary site of carrier-mediated bile salt and vitamin B12 absorption. Loose intercellular junctions contribute to significantly greater water and sodium fluxes in the jejunum than in the ileum. Tighter intercellular junctions and active transport of sodium chloride allow for significant fluid reabsorption and concentration of luminal content in the ileum (and colon). However, distinctions between jejunum and ileum are of clinical importance only if a significant length of bowel is resected. The ileocecal valve is thought to act as a “break” to the delivery of small bowel content into the cecum. It may also be a barrier for reflux of colonic content into the small bowel. However, ileal peristalsis probably is the main factor in those functions.
FIGURE 31-3 (A and B) The jejunum and ileum can be distinguished from one another by differences in luminal diameter, number of arterial arcades, and the presence or absence of fat encroaching on the gut wall. (Reproduced with permission from Agur AMR, Dalley AF, eds. Grant’s Atlas of Anatomy. 11th ed. Copyright Lippincott Williams & Wilkins; 2004.)
The luminal content of the proximal small bowel is of neutral pH and is relatively sterile, containing few bacteria. Most studies of the small bowel microflora have demonstrated increasing bacterial counts with distance away from the pylorus.5 The proximal small bowel flora resembles the gastric flora. The jejunum and proximal ileum contain gram-positive and gram-negative organisms at 104–105 cfu/mL. The bacterial concentration in the distal ileum rises to 105–108 in the ileum. There is also a higher number of anaerobic species in the ileum. This increase in bacterial load in the ileum is thought to contribute to an increased risk of infection with full-thickness injury in the distal small bowel versus the proximal small bowel.
Mechanism of Injury/Pathophysiology
Blunt injuries to the stomach and small bowel are infrequently encountered. Injuries include contusions, intramural hematomas, lacerations, full-thickness perforations, and mesenteric avulsions. In the East Association for the Surgery of Trauma (EAST) multi-institutional study, HVI was noted in only 1.2% of over 225,000 admissions during the 2-year study period.6 Most HVIs in this study were hematomas and several tears. Perforated small bowel injury accounted for less than 0.1% of blunt trauma admissions. Full-thickness perforations of the small bowel (unspecified site) and jejunum/ileum each accounted for 20–25% of all full-thickness HVIs. Gastric perforations following blunt trauma are rare and accounted for only 2.1% of the total HVI in the EAST study.
Most gastric injuries are related to pedestrian motor vehicle or high-speed motor vehicle crashes. The stomach is thick walled and relatively resistant to a blunt injury. However, when full after a recent meal, trauma to the left side of the body or inappropriate use of seat belts may contribute to rupture. Blunt gastric injuries include lacerations and full-thickness perforations, which most frequently involve the anterior gastric wall.7,8 Peritoneal signs and blood nasogastric tube aspirate are usually present and lead to early surgical intervention. Associated injuries are often severe because of the degree of force necessary to produce a gastric blowout.9 Associated injuries include liver, spleen, and pancreas, as well as injuries to the chest and head. Associated injuries are the main reason for the higher mortality rates for patients with blunt gastric rupture versus other HVIs.10–12
Small bowel perforation secondary to blunt abdominal trauma is uncommon, but thought to be increasing in incidence. Motor vehicle crashes are the most important mechanism for blunt intestinal trauma, followed by falls and bicycle accidents. Localized blows to the abdominal wall may also cause HVI. Mechanisms postulated for injury to the intestine to occur include: (1) crushing of bowel against the spine, (2) sheering of the bowel from its mesentery of a fixed point by sudden deceleration, and (3) bursting of a “pseudo-closed” loop of bowel owing to sudden increase in intraluminal pressure. More recently Cripps and Cooper have demonstrated experimentally the potential for small intestinal injury in high-velocity, low-momentum impacts that do not greatly compress the abdominal cavity.13 Earlier series had described proximal jejunum and distal ileum as sites more prone to injury due to their relative immobility. More recent reports have not shown this to be the case, as mid-jejunal perforations have been increasingly reported.
Although the use of seat belts alone or in combination with air bags is effective in reducing fatalities, there may be an increase in injuries associated with both proper and improper use of these devices. Garret and Brownstein first referred to the seat belt mark as ecchymoses across the abdominal wall that corresponds to the lap belt14 (Fig. 31-4). With the advent of the three-point restraint system, injuries may also involve the neck and chest. The “seat belt syndrome” now refers to intestinal injuries associated with lumbar fractures and abdominal or chest wall ecchymoses.14 Injury to other organs may occur and include the stomach and colon.
FIGURE 31-4 Patients with blunt intestinal injury sometimes have ecchymoses of the abdominal wall caused by restraint devices. The findings may be relatively subtle (A) or more severe (B). The presence of such ecchymoses does not always signify underlying blunt intestinal injury. By the same token, many patients with blunt intestinal injury do not have abdominal wall ecchymoses. The patient in (A) had a grade II injury while a grade IV injury (C) occurred in patient (B).
Anderson et al. reported a 4.38-fold increase in risk of small bowel injuries with lap/shoulder restraint use and a more than 10-fold increase in risk with lap belts alone, compared with no restraint use.15 Chandler et al. reported 112 patients involved in motor vehicle crashes.16 Sixty percent of patients were wearing a seat belt, and the remainder were unrestrained. There was no difference in the overall incidence of abdominal injury between belted and unbelted patients (15% vs. 10%, respectively). However, the incidence of small bowel perforation was significantly increased in patients with a seat belt versus no belt (6% vs. 2.2%, respectively). The presence of a seat belt sign (SBS) was associated with an even greater likelihood of abdominal injuries and small bowel perforation (64% and 21%, respectively).
More recently in the EAST multi-institutional study, the SBS was associated with a 4.7-fold increase in relative risk of small bowel perforation in patients following motor vehicle crashes.17 The second highest relative risk of small bowel perforation was the use of a seat belt without evidence of an abdominal seat belt mark (2.4-fold increase in relative risk). Small bowel injuries noted with the use of seat belt use include small bowel transections (usually in the proximal jejunum) as a manifestation of a deceleration injury, sheering or crushing injuries usually involving the terminal ileum and associated mesentery, and (blowout) perforations on the antimesenteric aspect of the bowel. This latter injury is felt to be due to an acute sudden increase in intraluminal pressure in a functionally closed loop of bowel. It is believed that when air bags are deployed in combination with a properly placed seat belt, there is a decrease in the incidence of abdominal injuries.
Children with an “SBS” may also have a higher rate of gastrointestinal injury. Sokolove et al. demonstrated that children with an SBS had a significantly greater risk of intra-abdominal injury, including gastrointestinal and pancreatic injuries.18 However, the increased risk of injury was only apparent in patients with abdominal pain or tenderness. In a study by Chidester et al., the SBS only had a sensitivity of 25% and a specificity of 85% for abdominal injury.19 Similar to the study by Sokolove, the presence of SBS with abdominal tenderness was more predictive of abdominal injury.
The association of a Chance-type fracture of the lumbar spine as a predictor of HVI is variably reported in the literature (Fig. 31-5). Anderson et al. reported 62.5% of 16 patients with Chance-type fractures had HVIs.15 Nine perforations occurred in the small bowel, and the remainder were in the colon. However, in the EAST multi-institutional trial with small bowel perforations there was no difference in incidence of Chance-type fracture in perforating or nonperforating small bowel injury patient groups versus patients without small bowel injury.17 The incidence of bowel perforations was quite low in all groups, and ranged between 2% and 3% of patients.
FIGURE 31-5 Some injuries to the lumbar spine are commonly caused by seat belts, and are frequently accompanied by associated blunt intestinal injury. Transversely oriented fractures through bone (A) are also sometimes known as Chance fractures. The mechanism of injury responsible for such fractures can cause soft-tissue disruption and dislocation in the same orientation as seen with Chance fracture (B).
In about 20% of patients with blunt intestinal perforations no other injuries are present. Other patients have significant extra-abdominal injuries with blunt injury as their sole intra-abdominal injury. Approximately 25% of patients with blunt intestinal injury have more than one injury requiring surgical intervention.20–22 Thus, in patients undergoing laparotomy for blunt intestinal rupture a complete evaluation for other injuries and a thorough laparotomy are mandated.
On rare occasions, patients may return to the hospital several days or weeks after blunt abdominal trauma with signs and symptoms of bowel obstruction. Contrast-enhanced CT of the abdomen performed at this time usually shows a thickened bowel loop, and narrowing of the lumen. This finding is due to intestinal stenosis resulting from mesenteric vascular injury. The stenosis is felt to be due to infarction resulting from the mesenteric injury rather then a direct injury to the intestine.23
Penetrating injuries to the stomach and small intestine are often more obvious. The anatomic location and space occupied by these organs make them the prime target following injury due to knives, gunshot wounds, shotgun wounds, and other piercing instruments. Of those with peritoneal penetration only 30% of patients with knife wounds have significant injuries requiring operation, whereas over 80% of patients who suffer gunshot wounds have injuries requiring surgical repair. Thus, most institutions employ selective observation of patients with knife wounds even with peritoneal penetration. The decision for operation is based on clinical signs of peritonitis. Many institutions apply a selective approach to shotgun wounds.24 For non-close-range shotgun wounds, operative intervention is based on the range of the blast and pellet distribution as well as an estimate of the number of pellets penetrating the peritoneal cavity. Patients with shrapnel wounds to the abdominal region may also be managed selectively, depending on findings from the physical exam and imaging studies.25
Blast injuries to the GI tract are the result of a “multidimensional injury” as four separate mechanisms may play a role.26 The primary blast injury results from an overpressure wave induced by the blast itself. Although primary blast injuries to the GI tract more commonly occur in the colon, the small bowel may also be affected. Exposure to extreme blast overpressure (which is invariably fatal) results in immediate lacerations of the bowel.
Nonfatal blast exposure may result in multiple contusions or intramural hematomas, which may evolve to full-thickness injury. The initial injury involves the mucosa–submucosa of the bowel wall; the presence of serosal injury is evidence of a transmural lesion at high risk of perforation. Because of the nature of this injury, there may be delay of 1–2 days, and rarely up to 14 days, before clinical symptoms occur. The overall incidence of bowel perforation is low (0.1–1.2%) but is increased with explosive amount, or when the victim is close to the center of the explosion or in an enclosed area.
Secondary blast injuries are caused by projectiles from the explosion that cause perforating injury to the victim. Tertiary blast injuries are the result of the generation of “blast winds” that propel the victim into rigid objects causing blunt injury. Quaternary injuries are the result of fire and heat generated by the explosion. Most injuries seen clinically are due to secondary or tertiary blast effects. Patients with penetrating torso injury or involving >4 body areas are at high risk for intraperitoneal injury.26 Patients with abdominal trauma after terror-related blast injury have a higher incidence of bowel injury (71.4%) and a lower incidence of solid organ injury (33%).26 Shrapnel is the leading cause of abdominal trauma in this setting.27 In the presence of peritoneal signs, the decision to perform surgery is easily made. Because bowel perforation may be delayed, careful observation is critical, even with negative initial image studies or diagnostic peritoneal lavage (DPL) results.
DIAGNOSIS
An accurate history of the traumatic event can help determine the potential for intra-abdominal injuries. In patients with a knife wound of the left thoracoabdominal region, diaphragmatic and gastric injuries are a primary concern. The small bowel is at risk of perforation following virtually any penetrating injury that violates the peritoneum. Evisceration of abdominal contents after abdominal stab wound is associated with significant intra-abdominal organ injury in 75% of patients even with no overt clinical signs that would mandate laparotomy. Certain patterns of blunt abdominal injury should alert the clinician as to the probability of gastric and small bowel perforations. Thus, a low threshold for laparotomy is appropriate in this setting. These include the use of seat belts, handle bar injury, and blows to the abdomen such as being kicked by a horse or other large animal. At the very minimum, individuals with an SBS should be admitted and observed with serial abdominal exams.
The incidence of small bowel injury in patients diagnosed with solid organ injuries by CT is variable. Nance et al. in a review of 3,089 patients with solid organ injury from the Pennsylvania Trauma Systems Foundation found 296 patients who had an HVI (9.6%).22 The frequency of HVI increased with the number of solid organs injured: 7.3% with one solid organ injury, 15.4% with two solid organ injuries, and 34.4% with three solid organs injured. More recently, Miller et al. reviewed the Memphis experience with nonoperative management of 803 hemodynamically stable patients with blunt liver or spleen injuries.21 Overall, the incidence of associated intra-abdominal injury was higher in the patients with liver injury at 5% as compared with 1.7% of patients with isolated splenic injury. Bowel injury was discovered in 11% of liver injury patients and no patients with isolated splenic injury. It is believed that the blunt force capable of producing liver injuries or multiple solid organs places the small bowel at increased risk for perforation and should arouse clinical suspicion for bowel injury.
Patients with penetrating gastric injuries usually present with significant peritoneal signs due to the peritoneal irritation from the intraperitoneal leakage of the low pH content of the stomach. Bloody nasogastric aspirate or free air demonstrated on an upright chest x-ray may be indicative of gastric injury but is neither completely sensitive nor specific for the presence of gastric injury.
In patients with obvious peritoneal penetration clinical findings following penetrating trauma to the small intestine may be minimal at first. This is because the luminal content of the small bowel has an almost neutral pH and is relatively sterile. Intestinal spill may also be relatively small, limiting the initial inflammatory response.
In the East multi-institutional trial of blunt small bowel injury, 1.2% of 227,972 blunt trauma admits were found to have a hollow viscous injury.17 A total of 72.5% of patients with perforating small bowel injury had abdominal tenderness; however, only 33.5% had peritoneal signs. Nonetheless, a careful physical exam by an experienced surgeon may discern the likelihood of intestinal perforation. In patients with an SBS on the abdominal wall, tenderness or guarding away from the seat belt should heighten the concern for the possibility of perforated small bowel injuries. Perforations of the stomach and small bowel are recognized by signs of peritoneal irritation: tenderness with guarding and rebound. Sensitivity of clinical examination to identify patients in need of operation exceeds 95% for stab wounds and gunshot wounds. In other studies, clinical examination of the abdomen has been shown to be unreliable in approximately 50% of blunt abdominal trauma patients.17 Significant limitations include patients with head injury and altered level of consciousness, intoxication due to drugs or alcohol, and spinal cord injury. The variable effect of hemoperitoneum from associated solid organ injuries and the presence of distracting injuries (e.g., pelvic fracture) in the multi-injured patients may also limit the clinical reliability of the findings on physical exam.
Laboratory studies including hematocrit, WBC, and serum amylase are not useful in the initial evaluation of patients with gastric and small intestinal injuries.17 In patients managed nonoperatively with solid organ injury or in patients with penetrating injuries undergoing serial clinical exams, unexplained tachycardia, hypotension, leucocytosis, an increase in serum amylase, or the development of a metabolic acidosis should arouse suspicion of a missed HVI.
A variety of diagnostic tests have been used to further evaluate the abdomen following blunt and penetrating injuries (see Chapters 15 and 16). DPL is very sensitive in detecting intraperitoneal injury but is now infrequently used. The most common peritoneal lavage finding with bowel injuries is gross blood.28 However, this may be due to associated solid organ or mesenteric injuries.
Peritoneal lavage with blood cell count has been used to diagnose HVI. However, Jacobs et al. found that a lavage white blood cell count >500/mm3 as the sole positive lavage criterion is a nonspecific indicator of intestinal perforation.29 It was suggested that sequential determinations of DPL and WBC may be useful in the diagnosis of intestinal perforation. Repeat lavage, diagnostic laparoscopy, or limited laparotomy in the patient already in the operating room for repair of other injuries may be prudent.
DPL amylase and alkaline phosphatase levels may also be useful in identifying HVIs. Jaffin et al. found an alkaline phosphate level >10 IU in the DPL effluent to have a specificity of 99.8% and a sensitivity of 94.7% in detecting small bowel and colonic injury.30 McAnena et al. used lavage amylase >20 IU/L and lavage alkaline phosphatase levels ≥3 IU/L as predictors of HVIs following blunt penetrating trauma.31 These values had a sensitivity of 54%, specificity of 48%, and a positive predictive value of 88% for significant abdominal injury.
The ability of DPL to detect hollow viscus perforation in the presence of hemoperitoneum secondary to solid organ injury may be improved by adjusting the positive criteria for WBC. Otomo et al. proposed a “positive” WBC criterion of WBC ≥ RBC/150 when peritoneal lavage was positive for hemoperitoneum.32 These criteria had a sensitivity of 96.6% and a specificity of 99.4% for intestinal injury when performed more than 3 hours after injury. Fang et al. used a “cell count ratio” in diagnosing hollow viscus perforation.33 The cell count ratio was defined as the ratio between white blood cell count and red blood cell count in the lavage fluid divided by the ratio of the same parameters in the peripheral blood. A cell count ratio of ≥1 predicted hollow viscus perforation with a specificity of 97% and a sensitivity of 100% when performed before 1.5 and 5 hours from the time of injury. The “lag time” between intestinal perforation and peritoneal white cell response was felt to account for the reliability of the “corrected” peritoneal lavage white blood cell counts calculated in these later two studies to detect hollow viscus perforations.
The Focused Assessment by Sonography for Trauma (FAST) has been a widely used test in the initial evaluation of suspected abdominal trauma. This is not as sensitive as DPL or CT in detecting stomach or small bowel injuries. This is most likely because of the relative inability of the FAST exam to pick up small amounts of free fluid typically found with isolated hollow viscus perforations. Rozycki et al. reported on 1,540 patients (1,227 with blunt injuries, 313 with penetrating injuries) who had FAST examinations performed as part of their initial assessment following injury.34 The sensitivity and specificity for detecting hemoperitoneum were 83.7% and 99.7%, respectively. However, there were 16 blunt abdominal trauma patients with false-negative ultrasound results. Three of these patients were subsequently found to have significant small bowel injuries. As an isolated finding, an additional patient had small bowel perforation and a ruptured bladder. Bowel injuries have been also missed by FAST examinations in patients with pelvic ring fractures. Detection of actual bowel injuries by FAST is unreliable.
CT scan is the most commonly used diagnostic modality in evaluating the abdomen in hemodynamically stable blunt trauma victims. It is occasionally used in evaluating hemodynamically stable patients with penetrating injuries particularly to the back and flank areas.
It is apparent that when blunt small bowel perforation is present, abdominal CT is usually abnormal.35–38 A number of CT findings are indicative or more often arouse suspicion for significant bowel and mesenteric blunt injuries. CT findings specific of bowel perforation include extraluminal oral contrast and discontinuity of hollow viscus wall. However, as oral contrast extravasation is noted in less than 10% of documented cases of small bowel perforation, it is not advocated for routine care at most centers (Fig. 31-6). CT findings suggestive of bowel injury include pneumoperitoneum, gas bubbles close to the bowel wall, thickened (>4–5 mm) bowel wall, bowel wall hematoma, and intraperitoneal free fluid without solid organ injury. Active contrast extravasation is a specific sign of mesenteric laceration, and mesenteric hematoma or fluid collection in the mesenteric folds is a CT finding suggestive of mesenteric injury (Fig. 31-7).
FIGURE 31-6 This CT was obtained a few hours after injury. The findings were highly suspicious of oral contrast extravasation (arrows). The patient was found to have rupture of the small intestine at surgery.
FIGURE 31-7 This CT was done shortly after injury. Multiple abnormalities are noted including thickened bowel and free fluid (arrows). However, a large amount of free air is the most notable finding. At operation a mid-jejunal perforation was found.
Malhatra et al. reviewed the Presley Regional Trauma Center experience with screening helical CT evaluation of blunt bowel and mesenteric injuries.38 One hundred of 8,112 scans were suspicious of blunt bowel/mesenteric injuries. There were 53 patients with bowel/mesentery injuries (true positive) and 47 without (false positive). The most common finding in both true-positive and false-positive groups was unexplained intraperitoneal fluid present in 74% and 79% of scans, respectively. Pneumoperitoneum and bowel wall thickening were much more common in true-positive scans. Multiple findings suspicious for bowel/mesenteric injury were seen in 57% of the true-positive scans but in only 17% of false-positive scans. The overall sensitivity and specificity of CT for bowel/mesenteric injury were 88.3% and 99.4%, respectively. The positive and negative predictor values were 53.0% and 99.9%, respectively.
In the EAST study, free fluid without solid organ injury had a 38.4% incidence of perforating small bowel injury.17 Even with the use of multidetector CT scans, free intraperitoneal fluid is the most common finding of blunt infected or mesenteric injury.36,39 It is important to semiquantify free fluid to help distinguish significance. Minimal fluid is defined as fluid in one anatomic region, and large amount of fluid is defined as fluid in multiple areas.40 Patients with minimal fluid can be followed by clinical exam or repeat CT imaging. Patients with larger amounts of fluid may be best served by DPL, diagnostic laparoscopy, or operative exploration. Of note, 12.2% of patients diagnosed with small bowel perforations in the EAST trial had a completely normal CT.17 Thus, if associated injuries do not mandate admission, a short period of observation may still be warranted, particularly if there is abdominal tenderness or the drug and/or alcohol screen is positive.41
In the initial report, Velmahos and colleagues demonstrated the safety of selective nonoperative management in over 1,800 patients with abdominal GSW.42 During observation, 80 of 792 patients selected for nonoperative management developed symptoms and underwent laparotomy, which was therapeutic in 57. In a follow-up study on 100 patients with nontangential abdominal GSW, the sensitivity and specificity of CT scanning were 70.5% and 90%, respectively.42 The two false-negative scans involved HVIs, which were repaired without complication. The most sensitive indicator for laparotomy for HVI from penetrating trauma is extension of the wound track, to or near a hollow viscus43 (Fig. 31-8). Therefore, oral or rectal contrast may not be necessary when used to evaluate patients with penetrating abdominal injuries.44 A recent meta-analysis by Goodman et al. demonstrated CT scanning in hemodynamically stable patients who had high (>90%) specificity, negative predictive value, accuracy, and slight lower positive predictive value.45
FIGURE 31-8 A single contact CT in a patient with tangential GSN shows the missile tract in close proximity to bowel with air around bowel loops.
Diagnostic laparoscopy is occasionally helpful in avoiding laparotomy in hemodynamically stable patients with penetrating thoracoabdominal trauma.46 Indications include penetrating left thoracoabdominal trauma with suspected diaphragmatic injuries. On some occasions small diaphragmatic tears and even gastric perforations may be repaired using laparoscopic techniques. Penetrating injuries to the anterior right thoracoabdominal area and tangential gunshot wounds to the abdomen may also be evaluated laparoscopically. Indications for diagnostic laparoscopy are less certain for patients with blunt intestinal trauma. A major limitation cited with diagnostic laparoscopy is in the relative inability to detect hollow viscus perforations. Earlier reports demonstrate an excessively high rate of missed injuries.46 Kawahara et al. recently demonstrated their experience in 75 hemodynamically stable patients with suspected abdominal injuries.47
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
