Classic teaching is that the majority (about 90%) of gunshot wounds to the abdomen penetrate the peritoneum.⁷ However, recent studies looking at nonoperative management show that a larger number of nontangential wounds do not penetrate. If a patient is initially stable and peritoneal signs are absent, the rate is probably closer to about 40%; however, abdominal gunshot wounds associated with peritonitis or instability have penetrated the peritoneum.^8,⁹ The majority of wounds that penetrate the peritoneum require laparotomy for repair. The most commonly injured organs are the small bowel, colon, and liver, followed by vascular structures, the stomach, and the kidneys (Box 80.1).

Stab wounds, as opposed to gunshot wounds, tend to follow the track of the wound and have more predictability. Approximately one fourth to one third of anterior abdominal stab wounds penetrate the peritoneum. Of those that do penetrate, about one third cause intraabdominal injury that requires operative repair. In addition to wounds involving the abdominal cavity, injuries to the thoracic cavity must also be considered in patients with thoracoabdominal stab wounds or any gunshot wound.

Physical examination often plays a major role in the management of patients with penetrating abdominal trauma, especially those who are hemodynamically stable. Serial examinations are a common and time-tested management strategy for low-velocity wounds. Studies show that it is an effective approach and that delay in diagnosis, if less than 24 hours, does not lead to a significant increase in complications. Furthermore, the decrease in morbidity and cost of nontherapeutic laparotomy is considerable.¹⁰^–¹² In fact, in some centers, even patients with evisceration (especially omentum alone) without peritonitis are observed successfully after replacement of the eviscerated peritoneal contents, although such management remains controversial.¹³^–¹⁵

Differential Diagnosis and Medical Decision Making

Examination may be used as a sole modality or in conjunction with other modalities. In patients with an anterior stab wound, local wound exploration can be a valuable diagnostic aid and should shorten a prolonged evaluation or observation if the results are negative. Its utility is dependent on the wound’s mechanism and location. Stab wounds involving the anterior aspect of the abdomen are well suited for local wound exploration because many do not penetrate the fascia and can easily be visualized. Back, flank, and thoracoabdominal wounds do not allow clear delineation of the wound track and extent and are not candidates for this procedure. Likewise, this is a poor option for gunshot wounds or wounds from a sharp pinpoint instrument (e.g., ice pick).

Over the last decade, patients with gunshot wounds have also been managed by serial observation, though generally in conjunction with some other modality, usually CT. The length of observation varies by institution. Standard practice is usually a 24-hour observation period, but recent data suggest that all important injuries will become apparent within a 12-hour period.¹⁶ Until more data are available or a clear institutional protocol has been established, approximately 24 hours of observation seems appropriate.

Diagnostic Modalities

Plain Radiography

Although plain abdominal radiography rarely adds to the evaluation of patients with blunt abdominal trauma, in cases of penetrating trauma with projectiles or retained foreign bodies it allows one to account for bullets, shrapnel, and foreign bodies. If all foreign bodies are not accounted for, one must consider the possibility that the foreign body is intraluminal or intravascular and a potential source of emboli. In patients with stab wounds or injuries without a foreign body, radiography’s utility lies in ruling out a broken instrument in the abdomen or in evaluating a large amount of free air consistent with a hollow viscus injury.

In thoracoabdominal trauma, chest radiography can identify the presence of hemothorax, pneumothorax, and possibly diaphragmatic injury. Unfortunately, its sensitivity for pneumothorax is variable, and at least two separate radiographs 4 to 6 hours apart must be taken to exclude clinically significant pneumothorax. Diaphragmatic injuries are seen clearly on plain radiographs only about 30% of the time. They cannot be excluded on the basis of normal radiographic findings.

Ultrasonography

Focused abdominal sonography for trauma (FAST) is a useful test if positive (i.e., free fluid is found during the examination). Several authors have shown specificity and positive predictive values in the low 90th percentile for therapeutic laparotomy if ultrasound reveals free fluid. A positive ultrasound finding after penetrating abdominal trauma should lead to exploration by either laparotomy or laparoscopy. Unfortunately, although its specificity is very high, its sensitivity ranges from 21% to 70%, which is not acceptable to rule out an injury requiring laparotomy.¹⁷^–²¹ Therefore, FAST findings negative for free fluid should be considered indeterminate, and further observation or testing is required. FAST should be able, in trained hands, to rule out significant pericardial effusion after thoracoabdominal trauma.

Computed Tomography

The expanded use of CT is a major change in the initial evaluation of patients with penetrating trauma in the past decade. In the past, use of CT had been limited in patients with penetrating abdominal trauma because of the high incidence of bowel injury and its lack of sensitivity in diagnosing bowel and mesenteric injuries, as well as rents in the diaphragm. The newest-generation CT scanners (i.e., multidetector scanners), as well as increased familiarity with their use, have markedly improved resolution and diagnostic capabilities. It is generally agreed that CT scanning of stab wounds in stable patients without the need for immediate laparotomy is a useful approach and can obviate admission when the wound is found to be superficial. In addition, it may reveal the path of a knife, identify or rule out peritoneal violation, and show with increasing sensitivity signs of hollow viscus perforation (free intraperitoneal air, unexplained free fluid, or bowel edema). These signs remain excellent in diagnosing solid organ injury. In addition, CT may show a “contrast blush,” a sign of active bleeding or false aneurysms in patients with solid organ injuries, and may establish whether early laparotomy or angiographic intervention is warranted.²² Although the negative predictive value of the need for operative intervention is high, patients with peritoneal penetration and no other clear operative requirements still merit an overnight observation period. In the case of gunshot wounds the literature is a bit less clear, but it generally shows that for tangential wounds in a stable patient, CT is excellent for ruling out abdominal penetration.

Even though it has not been shown to have high enough sensitivity to rule out diaphragmatic injuries, CT has improved and may one day be useful for this role.²¹ CT is now accepted for use in stable patients with penetrating flank trauma. “Triple-contrast” (intravenous, oral, rectal) CT has been found to be highly sensitive in diagnosing injuries to retroperitoneal structures, including bowel and renal injuries. At this time, however, its sensitivity is too low to fully exclude a bowel injury, and a negative CT scan should be followed by a period of observation, usually 24 hours. The one caveat is that it should be clear on CT that the wound track is superficial and that no intraperitoneal or retroperitoneal penetration has occurred.

Diagnostic Peritoneal Lavage

Though not practiced as commonly as in the past, diagnostic peritoneal lavage (DPL) is still a useful and acceptable screening tool for penetrating abdominal trauma. In unstable patients with blunt trauma, it has been supplanted by ultrasonography. With penetrating trauma, its major benefit is its high sensitivity in screening for intraabdominal penetration and injury to abdominal structures. The main drawback, in addition to a small but real number of complications, is its lack of specificity. That is, DPL tends to diagnose injuries that may be treated by observation alone. For this reason, DPL is sometimes used in patients with penetrating trauma in conjunction with less invasive (in comparison with laparotomy) procedures such as laparoscopy. It is still not entirely clear where it best fits in the overall algorithm and is probably best reserved for patients with equivocal findings on examination or concern for diaphragmatic injuries. One recent multicenter study looking at the various diagnostic options for penetrating abdominal trauma found that in stable patients, DPL did not perform markedly better than admission and serial examination.¹⁷ In patients with penetrating trauma who are hemodynamically unstable, it is not usually required to confirm what is already a high pretest probability that surgery is required. It may be of use, however, in patients with other possible causes of instability, especially in those with thoracoabdominal trauma, although it does not screen for retroperitoneal injuries.

A grossly bloody lavage effluent (10 mL of blood on initial aspiration) is always considered positive, although even with this finding, some surgeons will not operate immediately. If the effluent is not bloody on initial aspiration, 1 L of normal saline is instilled and then drained, and the effluent is sent to the laboratory for analysis. In patients with penetrating anterior abdominal wounds, lavage is always considered positive if the aspirate contains food particles, bile, or urine.

Cell counts tend to be a bit more controversial in that no cell count value is universally accepted to be indicative of a positive lavage result. If one is looking for penetration only, which may be useful when assessing gunshot wounds or diaphragm injuries, some institutions use as little as 1000/mm³; a more common and specific value, however, is 10,000/mm³. Although some centers use the cell count as an indication for laparotomy, others consider it complementary to observation or use CT or laparoscopy, depending on the wound. When assessing anterior abdominal wounds for injury and not just penetration, many institutions consider a cell count between 50,000/mm³ and 100,000/mm³ to be positive; a white blood cell count greater than 500/mm³ has also been used. However, these values do not seem to be particularly sensitive or specific.²³ Alkaline phosphatase and amylase lavage effluent levels have been advocated, but it is not clear that they add much to the standard criteria.²⁴

Laparoscopy

Minimally invasive laparoscopic surgery has gained general acceptance as a diagnostic and therapeutic modality in several circumstances. Although the specific injury may not always be identified, laparoscopy appears to be highly specific in identifying the need for therapeutic laparotomy. More important, in many instances it can rule out the need for laparotomy completely with a high degree of sensitivity. Its most accepted use is for low-velocity wounds when diaphragmatic injury (i.e., wounds involving the left thoracoabdominal area) is a possibility. It is also a reasonable tool to use in stable patients with anterior abdominal wounds to screen for peritoneal penetration and the need for laparotomy. The overall benefit, however, in this group of patient, over less invasive approaches such as serial examination, has not been well proven. It may be used in the operating room as an initial intervention in a stable patient with penetrating abdominal trauma and equivocal findings on physical examination. It allows a survey of the abdominal contents, repair of minor injuries, and possible avoidance of laparotomy with its associated complications. Its major drawbacks are its relative invasiveness, need for general anesthesia, and cost, as well as some literature showing suboptimal sensitivity for hollow viscus injury.²⁵^–²⁷

Diagnostic Testing for Specific Injuries

Other Injuries

Isolated Bowel Injuries

Patients with bowel injury alone or injury to the liver or spleen without significant bleeding may have minimal symptoms initially. Given that hollow viscus injury is much more common with penetrating trauma than with blunt trauma, the EP must search for these sometimes subtle or delayed findings. Tenderness at the wound site is normal, but peritoneal findings such as diffuse tenderness and muscular rigidity, regardless of stability, are generally indicative of the need for operative intervention.²⁸ These findings may become apparent several hours into a patient’s course if the injury involves rupture of a hollow viscus. Over time, intraperitoneal inflammation increases, stimulates somatic pain fibers, and becomes evident clinically.

Diaphragmatic Injuries

Diaphragmatic injuries may have dramatic findings but can also be occult and be accompanied by symptoms or even fatal complications years after the initial injury. Any patient with a thoracoabdominal injury may have a diaphragmatic injury. The rate of diaphragmatic injury varies from 7% to 42% in patients with penetrating thoracoabdominal trauma. The diagnostic rate varies with the aggressiveness of evaluation; if it is not looked for, many diaphragmatic injuries will be missed. It tends to be highest in patients with left costal penetration.^25,^26,²⁹ Most left-sided tears can be repaired surgically. A small right-sided tear may not need to be repaired because herniation is much less likely on this side as a result of protection by the liver. In cases of potential left diaphragmatic injury, laparoscopy or DPL is a reasonable addition to the diagnostic approach.

Treatment

Most interventions and procedures in patients with penetrating trauma (Table 80.1) are aimed at first determining whether intraabdominal penetration occurred and, if so, deciding whether the patient can avoid a nontherapeutic laparotomy. The options are varied and may be performed alone or in series depending on the test, findings, and local custom. Although mandatory exploration was once the paradigm for penetrating injuries, knowledge of significant morbidity from negative laparotomy has pushed surgeons toward a much more conservative approach to management.^22,^30,³¹

Table 80.1 Options for Evaluation

DIAGNOSTIC MODALITIES	ADVANTAGES	DISADVANTAGES
Local wound exploration	Inexpensive bedside test; if negative, can discharge the patient	Operator dependent; may be inconclusive; not good for gunshot wounds or impalement
Ultrasonography	Inexpensive bedside test; high positive predictive value (90%) for therapeutic laparotomy	Operator dependent; poor sensitivity for bowel injury; if negative, cannot exclude injury
Diagnostic peritoneal lavage	Highly sensitive, inexpensive; can diagnose small bowel and diaphragm injuries	Poor specificity; up to 25% negative laparotomies when using lower limits of red blood cell counts
Computed tomography	Excellent for solid organ injuries; can often show lack of peritoneal penetration and obviate the need for observation; test of choice for flank and back injuries because it shows the retroperitoneal structures	Expensive; requires radiation; variable sensitivity for bowel and diaphragm injuries; unless penetration of the peritoneum is clearly excluded, observation is required afterward
Laparoscopy	Test of choice for left-sided thoracoabdominal wounds; can exclude peritoneal penetration and also screen for more serious injury; can be used for repair in selected cases	Expensive; associated complications; requires operating room

Exploration requires aseptic technique, good overhead lighting, and local lidocaine and epinephrine anesthesia. Inserting a digit or cotton swab into the wound is not an acceptable alternative. Obese or noncooperative patients and those with abdominal scarring from previous operations are not optimal candidates. Paralleling the natural skin lines, the wound is enlarged as necessary so that the posterior fascia may be evaluated. If penetration of the anterior fascia has occurred or if the wound exploration is inconclusive, the wound is considered intraperitoneal and must be evaluated further by more invasive methods or by observation and serial examinations. A recent multicenter observational study suggested that serial examinations are a safe option. Operative evaluation based solely on penetration would lead to a high number of nontherapeutic laparotomies (57%).¹⁷ If the fascia is clearly intact, the wound can be irrigated and closed by primary intention if clean and the patient discharged. Alternatively, the initial wound can be closed by secondary intention or delayed primary closure.²³

Fluid Resuscitation

Resuscitation of hypotensive patients remains a widely discussed and controversial area of treatment. The advanced trauma life support standard treatment is to administer 2 L of intravenous fluid followed by transfusion of blood; however, some EPs and surgeons prefer to limit resuscitation until definitive control is obtained ^32,³³ because of concerns that elevation in blood pressure through aggressive administration of fluids may disrupt clots and clotting factors, as well as evidence in multiple animal studies that less aggressive resuscitation or “permissive hypotension” leads to improved outcomes.

Two well-done large prospective trials have been carried out in humans, a study of penetrating torso injuries by Bickell et al. and a study of a combination of blunt and penetrating injuries by Dutton et al.^32,³⁴ The Bickell trial enrolled 598 patients with penetrating trauma and a systolic blood pressure of 90 mm Hg or lower. The authors found a trend toward lower mortality and morbidity in the minimally resuscitated patients, although the “appropriately resuscitated” patients—or those with standard high fluid volume—had blood pressure values that were significantly higher after resuscitation but before surgery. The Dutton study³² of approximately 100 patients compared the outcomes of actively bleeding patients treated via the standard resuscitation protocol with the outcomes of patients treated with a “hypotensive resuscitation” protocol consisting of a target systolic blood pressure of 70 mm Hg. Neither the standard group nor the hypotensive group showed improvements in mortality or worse outcomes; however, there was difficulty reaching the target blood pressures. In the authors’ judgment, limiting resuscitation with normal saline and blood to reach a target systolic blood pressure of around 90 mm Hg appears to be reasonable. Moving more rapidly to blood replacement rather than excessive saline is probably a reasonable approach also. In patients with massive blood loss, the current literature suggests blood replacement protocols with increased use of platelets and clotting factors, although the exact ratios remain unclear and somewhat controversial.^35,³⁶ It may not be feasible, appropriate, or beneficial to allow the blood pressure to decrease much lower.