Abdominal trauma [link]
Initial assessment [link]
Resuscitation room investigations [link]
Investigations outside the resuscitation room [link]
The trauma laparotomy and damage control [link]
Selective non-operative management [link]
Critical decision making in abdominal trauma [link]
Further reading [link]
The abdominal cavity extends from the inguinal ligaments below, to the costal margins above, with the domes of the diaphragm rising to the level of the 4th or 5th intercostal space in full expiration—the anterior axillary lines represent its lateral extremes. The flanks lie between the anterior and posterior axillary lines on both sides, and the back between the two posterior axillary lines. The abdominal viscera are distributed between the intraperitoneal and retroperitoneal compartments, both of which extend into the pelvis.
This contains the stomach, liver, spleen, 1st part of duodenum, the small bowel from jejunum downwards, the transverse and sigmoid colon, as well as the gynaecological organs in the female pelvis. The upper viscera lie under the protection of the lower rib cage but are susceptible to injury to the lower thoracic cage—the ‘intrathoracic abdomen’.
Lying posterior to the intraperitoneal cavity, the retroperitoneum is bounded behind by the thick musculature of the back and flanks; it is usually subdivided into four regions. The central zone contains the aorta, vena cava, pancreas and 2nd, 3rd and 4th parts of the duodenum. The kidneys with the adrenals and ureters, ascending and descending colon lie in the respective left or right lateral zones, and the pelvic retroperitoneum contains the rectum, bladder and iliac vessels.
Mechanisms of injury
Abdominal trauma may result from either blunt or penetrating injury. The majority of abdominal trauma in the UK is a result of blunt injury, most commonly from motor vehicle collisions (MVCs); the penetrating trauma that does occur is largely from stab wounds, with gunshot wounds remaining uncommon in British hospitals. In urban America and South Africa, up to 35% of trauma admissions are from penetrating injury, with more gunshot wounds than stabbings. Penetrating injury may also result from MVCs, domestic and industrial accidents. Blast injury represents a special mechanism of injury for both blunt and penetrating abdominal injury and is considered in detail in Chapter 24.
Injuries to both solid and hollow abdominal viscera may be graded, the most commonly used system being the one devised by the American Association for the Surgery of Trauma (AAST). This largely relies on the radiological appearances of the injury, with grade increasing with size of laceration, subcapsular haematoma, or involvement of the vascular pedicle in solid organ injury and degree of tissue loss in hollow organ damage.
Direct compression by a lateral or antero-posterior force will crush immobile viscera against the unyielding restraints of the abdominal cavity. Those organs with strong peritoneal attachments, such as the liver and spleen, as well as the duodenojejunal (DJ) flexure, are prone to this form of injury as are the retroperitoneal viscera. Direct rupture can occur with associated massive haemorrhage.
Rotational or deceleration forces applied to the abdomen result in differential movement of its viscera, with areas of relative fixity becoming the points of force concentration. The insertion points of blood vessels into viscera typically act as stress risers and these are commonly avulsed, leading to significant haemorrhage, as well as potential devitalization of distal parenchyma.
Acute compression of the abdominal cavity generates a sudden rise in intra-abdominal pressure and pressure within the lumen of hollow organs, which if sufficient will cause bursting. The oesophagogastric junction is particularly prone to this mechanism of injury, and most diaphragmatic ruptures occur in this manner as the increased abdominal pressure decompresses into the thorax.
Pathophysiology of penetrating injury
The pathophysiology of penetrating injury is dependant on the degree of energy transfer. Stab wounds create injury confined to the wound track, unless the stab severs neurovascular structures causing downstream ischaemic damage. The peritoneum is violated in only about half of abdominal stab wounds and only about one-half of these injure abdominal viscera. The mechanisms of ballistic injury are described in detail in Chapter 24, but in brief are caused by an energized projectile passing through tissue and being retarded by the drag exerted by the tissues. This transfers energy from the missile to the tissues and this energy performs work. There is a high pressure region in front of the missile which creates a permanent track analogous to that created by a knife wound; in addition, as the energy transfer increases, so does the degree of radial dissipation of energy in the form of a low pressure shear wave which pushes the tissue away from the permanent track in a temporary cavity, which may achieve a diameter of 30 times that of the missile. The temporary cavity is at subatmospheric pressure and sucks debris and contamination into the cavity, which is dissipated through the tissues as the cavity collapses.
Penetrating abdominal trauma may exist in the absence of external abdominal injury as the cavity may be breached transdiaphragmatically from the thorax, or from below through the buttocks, perineum, and groins.
Solid organ injury
The solid organs most frequently injured in blunt abdominal trauma are the spleen, liver, kidneys, and pancreas; the liver is more susceptible than the spleen to penetrating injury.
Diaphragmatic injury in blunt trauma is usually due to bursting as the abdominal pressure rises acutely and decompresses into the thorax; the left side is affected three times more frequently than the right, as the pressure is absorbed to some degree by the liver. Abdominal viscera may follow as a diaphragmatic hernia either immediately or many years later (Fig. 12.1).
The incidence of traumatic diaphragmatic injury is approximately 0.63%, two-thirds of which are in penetrating injury. In stab wounds the direction of the wound track may suggest a potential diaphragmatic injury, but such prediction is impossible after a gunshot wound. The diaphragmatic injury may occur in the opposite direction with the penetrating agent traversing from thorax into abdomen. Diaphragmatic injuries generally occur in association with other injuries in both blunt and penetrating trauma, and are a marker of increased injury severity. Nearly half of all diaphragmatic injuries, irrespective of mechanism, are associated with liver injuries and haemopneumothoraces.
The liver remains the most commonly injured organ in abdominal trauma, even though it is shielded by the lower right rib cage. In blunt trauma, compression and shearing are the predominant mechanisms of injury. The liver is covered in a fibrous capsule and attached to the abdominal wall by the triangular, coronary and falciform ligaments. When compressed, commonly against the lower ribcage, the liver cannot easily move out of the way and the parenchyma is lacerated.
If the overlying fibrous capsule remains intact then haemorrhage is contained as a subcapsular haematoma, with the possibility of subsequent capsular rupture and profuse haemorrhage. More significant degrees of energy transfer may rupture the liver capsule and parenchyma resulting in free intraperitoneal bleeding. The greater the level of energy transfer the deeper the liver is injured with severe injuries crushing the central segments and caudate lobe. In shear injuries, which may also be caused by compression, as the liver attempts to move out of the way of the compressive force, the attachment of the supporting ligaments to the capsule are disrupted, tearing the parenchyma attached beneath. In severe injury the entire porta hepatis may be avulsed.
Stab wounds cause injury only along the path of the implement and unless a large vessel or biliary radical is transected, usually have low clinical impact. Ballistic wounding of the liver, however, has the potential for massive disruption. Low energy transfer ballistic wounds are largely akin to stab wounds in that the damage caused is largely confined to the direct tissue track. As energy transfer increases, so does the degree of temporary cavitation; the liver parenchyma is relatively non-elastic and accepts the stretch of cavitation poorly, hence, hepatic tissue tends to disrupt, rather than stretch. In addition, the fibrous capsule resists elastic expansion of the cavity compounding the tissue disruption—high energy transfer gunshot injury to the liver results in significant tissue damage.
Injuries to the portal triad are rare, but more common after penetrating trauma. Injuries to the extrahepatic biliary tree are extremely rare; approximately one-third are from blunt injury. The mortality rate is approximately 50%, largely due to vascular injury and this increases to 99% if both portal vein and hepatic artery are injured.
In terms of trauma, the spleen bears many resemblances to the liver. It resides high in the abdominal cavity underneath the overlying ribcage, is surrounded by a capsule, and is anchored to the abdominal wall by a serious of fascial condensations referred to as the splenic ligaments. It is thus susceptible to injury from similar mechanisms. However, the capsule is more friable and tears more easily (Fig. 12.2).
Despite its relatively protected position inside the rib cage, the spleen is commonly injured in blunt abdominal trauma, most frequently from MVCs, although sporting injuries, iatrogenic damage, and spontaneous rupture are well recognized. The vascular pedicle of the spleen is closely related to the tail of the pancreas and injuries involving the pedicle invariably involve the distal pancreas as well. Blunt splenic injury may generate parenchymal lacerations or subcapsular haematomas (Fig. 12.3); injury with capsular disruption leads to profuse free intraperitoneal haemorrhage. Penetrating splenic injury is less common than for the liver as the target is smaller and relatively more mobile but the pathological effects are similar.
Relatively protected in the retroperitoneum with the renal pedicle as the only point of fixation, the kidneys are normally only damaged in severe trauma (renal injury occurs in approximately 10% of all cases of abdominal trauma) (Fig. 12.4). Compression injuries require great force because of the protection offered by the posterior abdominal wall musculature, but the kidneys are susceptible to shearing of the pedicle in deceleration injuries.
The closed space of the retroperitoneum may tamponade the haemorrhage from significant renal injury. Blunt mechanisms predominate, even in gunshot-rich countries such as the USA. Injury rates are increased if the kidneys are abnormal, such as with cystic kidney disease, renal malignancy, and horseshoe kidneys.
The pancreas traverses the posterior abdominal wall from the duodenum to the spleen and, in contrast to the other abdominal solid viscera, is not encapsulated. Blunt pancreatic injuries are usually compression injuries as unrestrained vehicle drivers strike their torsos against the steering wheel, crushing the pancreas, in the region of its neck, against the vertebral column (Fig. 12.5). In children, up to 75% of pancreatic injuries result from bicycle handlebars compressing the epigastrium; the duodenum is also often injured in this manner.
Blunt pancreatic injury requires a significant degree of force and there will usually be associated visceral injury. Parenchymal disruption by compression may also injure the pancreatic duct, which markedly increases morbidity and mortality, as the cycle of pancreatic enzyme activation and autodigestion accompanied by a systemic inflammatory response is unlikely to abate spontaneously. Pancreatic injury in isolation occurs in less than 10% of cases with a mean of four associated injuries. Pancreatic trauma, whether blunt or penetrating, has a high mortality. Large institutional series describe death rates of 3–32% in penetrating injury with that from shotguns and gunshots far outweighing stab-related deaths. Blunt injury has a mortality rate of 17–21%. Most deaths occur within 48 h of injury from haemorrhage or associated injuries.
The gastrointestinal and urinary tracts comprise the abdominal hollow viscera. The majority of the gut is protected from injury by its mobility, which suggests that it is vulnerable at points of fixed attachment such as the DJ flexure, ileocaecal junction, and the retroperitoneal segments. The gut, when injured, suffers either from direct trauma, particularly shearing at the fixed points, or injury to the mesentery and disruption of the blood supply giving rise to segmental ischaemia (Fig. 12.6).
Gastric injury is relatively uncommon with direct penetrating injury commoner than blunt trauma. Blunt injury is often of the bursting type from acutely raised intra-abdominal pressure when the stomach is full, or shearing at the gastro-oesophageal junction. Penetrating injury to only one gastric wall is uncommon and surgery must inspect both anterior and posterior walls of the stomach. Blunt gastric perforations usually require significant force, are associated with multiple injuries, and have an increased mortality when compared with other gastrointestinal perforations.
Isolated duodenal injury is uncommon due to its proximity to other organs and the abdominal vessels. Epigastric compression (e.g. from handlebar injury) classically compresses the 2nd and 3rd parts of the duodenum against the vertebral column. If direct rupture does not occur, a mural haematoma may ensue, which increases in size over time and typically may present as epigastric pain and vomiting 2 or 3 days after injury as the haematoma obstructs the duodenal lumen. Shear injuries at the DJ flexure are common.
Small intestine and mesentery
The small intestine is injured by shear at the DJ and ileocaecal junctions and, occasionally, at other points fixed by congenital bands or adhesions. Direct injury may generate a mural haematoma, which if large will perforate immediately, but smaller lesions may gradually necrose the intestinal wall perforating up to 2 weeks after initial injury. Small bowel injury also occurs as a result of interruption of its blood supply because of mesenteric injury from either compression, shearing, or penetrating injury. If the ensuing ischaemia is not sufficient to cause acute perforation, segmental stenosis and later obstruction may occur. Mesenteric haematomas can extrinsically compress structures including the blood supply to their segments of gut. Free intraperitoneal bleeds from mesenteric vessels are often profuse as there is little ability for tamponade to limit the bleeding.
Blunt colonic trauma accounts for only about 5% of colon injuries. Deceleration causes shearing at the junctions of the intra- and retroperitoneal portions; the whole colon is susceptible to compression, and burst injuries and mural contusions follow the same pattern as in the small bowel. Colonic perforation may also occasionally occur as a result of extraperitoneal passage of energized missiles by virtue of the shear wave generated by temporary cavitation. Damage to the colonic blood supply is less common, but again may cause ischaemic perforation or delayed ischaemic stricturing. Colonic injuries are present in over one-third of patients with penetrating abdominal trauma with the transverse colon being most commonly affected; multiple colonic penetration occurs in a quarter of cases. Stab wounds tend to produce through-and-through wounds rather than injuring only one wall of the colon. Iatrogenic colonic injury during endoscopic examination occurs in approximately 0.1% of colonoscopies.
The rectum is protected by the bony pelvis, which limits its susceptibility to compression and shear injuries. Blunt rectal injury is rare. It is prone to penetrating injury by bone fragments generated by pelvic crush injury and by any pelvic or gluteal gunshot. It may also be injured by anal insertion of foreign bodies (Fig. 12.7a&b).
Both ureters are well protected in the retroperitoneum. Blunt trauma accounts for only 10% of ureteric injuries (and ureteric injury accounts for only 3% of all urinary tract trauma aside from iatrogenic surgical damage), but such injuries should be considered in children who have suffered spinal hyper-extension trauma, as this may avulse the ureters at the pelvi-ureteric junction. Pelvic injuries involving the ureters are usually associated with concurrent iliac vessel injury and a high mortality rate. The ureters are at risk more from anterior than posterior stab wounds as the paravertebral muscles protect from behind, and injury in gunshot wounds is again associated with multi-organ injury (Fig. 12.8).
The bladder is an extraperitoneal organ and may be injured by blunt or penetrating injury. Blunt trauma, usually from MVCs, can shear the bladder at its attachments to the pelvis or pelvic compressive trauma will generate penetrating bone spicules, typically causing extraperitoneal injuries. Direct trauma to the dome of a full bladder, typically a kick to the lower abdomen in a drunken brawl, may cause the bladder to rupture intraperitoneally and generate a chemical peritonitis.
In the male, the external genitalia is susceptible to blunt and penetrating trauma. The commonest blunt injury is a penile fracture after forced bending of the erect penis during sexual intercourse. Blunt scrotal trauma has a high rate of testicular rupture. Penetrating penile or scrotal injury is associated with a high incidence of related injuries such as rectal penetration and femoral nerve and vessel injury.
The abdominal vasculature is considered in five separate, but contiguous regions. Zone 1 is the midline retroperitoneal structures, which are further subdivided into those above the transverse mesocolon [suprarenal aorta, coeliac axis, superior mesenteric vein (SMV), and artery (SMA) and proximal renal arteries] and those below (infrarenal aorta and infrahepatic vena cava); zone 2 is the upper lateral retroperitoneum containing the renal vessels; and zone 3 (pelvic retroperitoneum) contains the iliac vessels. The retrohepatic vena cava, hepatic artery and portal vein constitute the final area.
Abdominal vascular injury occurs predominantly from penetrating mechanisms as all the major vessels are relatively well protected from blunt mechanisms throughout their abdominal course. Rapid deceleration during an MVC generates shearing forces on the origins of small vessels and may avulse them—typically either proximal or distal intestinal branches of the SMA. Direct trauma will either cause rupture, which may be intra- or retroperitoneal, with mortality being much higher for free intraperitoneal bleeds or intimal injury giving rise to later thrombotic occlusion. A well-recognized injury is the seatbelt aortic injury, where direct abdominal crush raises an intimal flap in large vessels, such as the infrarenal aorta or superior mesenteric artery. Ninety-five per cent of abdominal vascular trauma has a penetrating aetiology and vascular injury can account for nearly one-third of all abdominal trauma in a civilian context although this figure is much lower (approximately 5%) in military conflicts. This difference is a combination of the widespread use of personal ballistic protection of the torso in the military (body armour) and the tendency for military wounds to be of much higher energy transfer. Such wounds within the abdomen result in significant cavitation and extensive damage and, if accompanied by a major vascular injury, are unlikely to survive to definitive hospital care. Abdominal vessel injury is rarely an isolated phenomenon and associated small bowel injuries are common.
Each organ and mechanism of injury have been described in isolation, but it is clear that many injuries occur as part of well recognized injury complexes, often with a specific aetiology.
Seat belt sign
Bruising of the anterior abdominal wall in the pattern and distribution of a restraining seatbelt is well recognized as a marker of potential underlying internal injury. This is true for three-point, shoulder and solitary lap belts, although the risk is greatest from a lap belt alone. There is a three fold increase in the risk of small bowel perforation if a seatbelt is worn (6 versus 2.2%) and a similar increased risk but substantially higher incidence (64 versus 21%) if a seatbelt sign was present or not. The commonest sites of injury are the proximal jejunum (deceleration injury), terminal ileum (shearing or crushing) and blowouts from sudden rises in intraluminal pressure.
The Chance fracture is a purely bony injury of the spinal column as a result of forced forward flexion—typically as a motor vehicle occupant is thrown forward against a lap belt in an MVC. This flexion-distraction injury typically occurs at the level of the thoracolumbar junction or first 2 lumbar vertebrae; in children the injury may be lower due a lower centre of gravity. This fracture pattern is associated with a high rate of intra-abdominal injuries, most commonly small bowel perforation—rates of up to 60% have been reported, but a recent multicentre study of 79 patients reported a 33% incidence intra-abdominal injury, most of which were small bowel perforations. The incidence is usually reported as being slightly higher in paediatric Chance fractures. Approximately three-quarters of children with seatbelt bruising and a Chance fracture will require therapeutic laparotomy.
Blunt pelvic fractures may be classified by the net direction of the applied force and the degree of haemorrhage is related to this. Lateral compression injuries of the pelvis are typically associated with significant injuries elsewhere (classically from an MVC side impact), but as they act to shorten the pelvic vasculature, haemorrhage is relatively modest. A similar effect is seen with vertical compression injuries, such as those suffered when jumping from a height. Anteroposterior (AP) compression, however, widens the pelvis and the hypogastric plexuses are frequently injured. High grades of AP compression may also disrupt the iliac vessels.
Figure 12.9 offers clues as to the likely concomitant injuries after pancreaticoduodenal trauma. The profusion of major vessels means that significant vascular damage is usually associated with injuries to the pancreas and duodenum, and this contributes to the significant morbidity and mortality previously described. There is a 40% chance of major vessel injury after pancreatic injury and 22% rate of aortocaval injury with duodenal wounds.
The presentation of abdominal trauma varies widely between haemodynamic stability and complete cardiovascular collapse with abdominal signs across the spectrum from normality to frank peritonism. Thus, assessment must be concise and timely, and should be repeated as often as necessary to ensure that nothing is missed.
Injury to the solid organs, blunt or penetrating, usually presents with hypotension from haemorrhage. Blood can be surprisingly non-irritant and so peritoneal signs may be absent initially, but tenderness in the respective upper quadrant is common. Lower rib fractures raise the suspicion of hepatic or splenic injury in blunt trauma. If the visceral injury is contained within the fibrous capsule then a subcapsular haematoma develops. Upper quadrant tenderness is present, but widespread peritoneal signs are generally absent as free rupture has not (yet) occurred. Significant intraperitoneal haemorrhage also occurs from damage to the mesenteric vessels and can be life-threatening, as there is nothing to tamponade the bleeding, especially if partial transection has occurred. Complete transection of any vessel normally results in retraction and contraction of the injured vessel with arrest of haemorrhage.
Different patterns of pain can offer clues as to the site of injury. Peri-umbilical pain arises embryologically from the mid-gut, which extends from the 2nd part of duodenum to distal transverse colon, while low abdominal pain arises from the hindgut. Pain between the shoulder blades is typical of diaphragmatic peritoneal irritation, although in cases of deceleration injury thoracic aortic transection can give similar pain. Epigastric pain boring through to the back is typical of pancreatic disease. Retroperitoneal injury requires a high degree of suspicion as the signs and symptoms may be subtle. Fullness, bruising, and tenderness in the flank raise the possibility of renal tract injury—haematuria offers confirmation, but not anatomical localization.
Abdominal trauma, particularly blunt injury, is rarely an isolated occurrence and a systematic approach to the whole patient is required. The traditional approach of primary and secondary survey places the abdominal assessment into the secondary survey, but abdominal injury may be the contributing aetiology of a persistent hypotension that must be addressed in the circulatory section of the primary survey or junctional vascular injury leading to catastrophic cardiovascular collapse, which necessitates intervention even before management of the airway.
This may be obtained from the patient, friends and relatives, bystanders, or the police and ambulance crews. Information about the mechanism of injury gives useful clues to the potential for intra-abdominal injuries, but should be viewed in the context of the clinical presentation. In blunt trauma, details of the mechanism of injury allow an estimation of the level of energy transfer—in MVCs this includes the speed of impact, and the use of seatbelts and airbags. Other features such as collapse of the steering column, distortion of the steering wheel or ejection from the vehicle suggest an increased degree of energy transfer. In penetrating trauma, knowledge of the weapon is useful, in particular the length and type of knife blade used, the number of potential stab wounds, and the angle from which the wounds were inflicted. In gunshot wounds, the type of weapon offers a guide to the potential energy transfer, but this should not detract from a thorough clinical assessment. The adage treat the wound not the weapon remains apposite. The number of shots and distance between gun and victim is useful, the latter particularly relevant in shotgun wounds where the available energy diminishes rapidly at ranges above 5 m—the majority of shotgun pellets do not penetrate skin at a range of greater than 12 m. An AMPLET history should be taken.
A single abdominal examination is unreliable in detecting intraperitoneal injury (sensitivity of approximately 50%), whereas serial examination by the same observer is much better at detecting the onset of subtle signs and changes. The examination must include inspection of the back and posterior torso, as it is an area where penetrating wounds are notoriously well hidden. The presence or absence of bowel sounds correlates poorly with the presence of intra-abdominal injury. Digital rectal and vaginal examination and inspection of the external genitalia are mandatory. It should be noted that, in certain circumstances, resuscitative laparotomy, or thoracotomy in the case of multi-cavity injury, may form part of the circulation section of the primary survey. Assessment of the abdomen should also include notice of other specific injuries known to be associated with intraperitoneal injury, such as seatbelt bruising and lumbar spine fractures.
Clinically evident peritonitis may be absent or difficult to distinguish from abdominal wall injury, or overlooked because of the distraction of other injuries. Assessment must include the haemodynamic status and response to fluid resuscitation, as this will guide the choice of further investigations.
Gastrointestinal injury is notoriously difficult to diagnose. Traditionally, small bowel perforation was diagnosed at laparotomy for visceral haemorrhage—an opportunity denied to the modern surgeon in the era of non-operative management of up to 90% of hepatosplenic injuries. In the absence of free intraperitoneal air, CT may only show a small amount of free fluid and is insensitive for the direct demonstration of small bowel injury. Delays in diagnosis of small bowel perforation are associated with increased mortality and morbidity. The remaining retroperitoneal gut gives rise to diffuse signs initially (non-specific anterior abdominal pain, and vague flank and back pain) and it may be the onset of sepsis that prompts consideration of retroperitoneal gastrointestinal injury. The overall impact of abdominal trauma is significant as it contributes to 20% of all trauma deaths by way of haemorrhage and sepsis.
In most instances of abdominal trauma it is the degree of haemodynamic instability and response to resuscitation that informs the decision making process as to the need and urgency of laparotomy and choice of investigation. Thus the abdominal assessment should pay attention to those parameters and adjuncts that may have been instigated in other parts of the primary and secondary surveys. Pulse and blood pressure are easily measured parameters and the level of haemorrhagic shock has been equated with the volume of blood loss. It should be remembered that such estimates may prove wildly inaccurate in the young and the old. Young fit adults may tolerate significant blood loss without a rise in pulse rate or fall in systolic blood pressure; similarly, the elderly may have little physiological reserve and cardiovascular collapse may ensue very early in the injury process. Urine output is a sensitive indicator of end-organ perfusion and a urinary catheter should be inserted early in the resuscitation. The urethral route should be used unless there is a contraindication suggestive of urethral injury. Markers of potential urethral injury are:
• Blood at the external urethral meatus.
• High riding prostate on digital rectal examination.
• Scrotal or perineal haematoma.
• Significant pelvic fracture.
In the adult, resuscitation should aim for a minimum urine output of 0.5 mL/kg/h (1–2 mL/kg/h in children). Arterial blood gas analysis may give evidence of worsening acidosis, falling oxygenation, or rising CO2 levels.
Where there is catastrophic cardiovascular collapse in the context of abdominal trauma—usually taken to mean imminent cardiorespiratory arrest from hypovolaemia—then immediate laparotomy to control haemorrhage is indicated. This should ideally be undertaken in the operating theatre. In such instances, it is important to remember to complete the remainder of the primary and secondary surveys as soon as is practicable so that other injuries are not overlooked.
Blood should be sent for laboratory analysis. The serum amylase is normal in nearly half of all cases of pancreatic injury on presentation but resampling every 6 h for the first 24 h may demonstrate a rising titre suggestive of pancreatic damage.
The abdominal assessment of a trauma victim seeks to answer three basic questions:
• Has the abdomen been injured?
• Does the patient need a laparotomy?
• How quickly is that laparotomy needed?
Any investigations performed in the resuscitation room should add to the clinician’s ability to answer those questions. More recently, a fourth question—does the patient require damage control surgery (DCS)?-has been added, the answer to which is based more on physiological data than investigations per se. Investigations that may be undertaken in the resuscitation room are to some degree dictated by the facilities and expertise available.
Antero-posterior chest and pelvic radiographs form part of the initial trauma assessment and should be performed in all patients with significant trauma. Aside from disclosing thoracic pathology, an erect or semi-erect chest X-ray may show free sub-diaphragmatic air indicative of intra-abdominal visceral perforation (Fig. 12.10); however, this is rarely possible in the trauma situation due to the necessity for spinal protection.
Fracture of the lower ribs increases the likelihood of injury to the spleen and liver, and gastric or small bowel shadows within the thorax—typically the left—suggest diaphragmatic rupture. Pelvic fractures on plain radiography should raise the suspicion of associated visceral injury. This can include avulsion of the prostatic urethra in males, extraperitoneal bladder rupture, or penetration of bladder or rectum by spicules of pelvic bone.
In ballistic penetrating trauma, plain abdominal X-ray may show radio-opaque foreign bodies and wounds should be marked on the surface (paperclips are ideal for this purpose). Free gas may be identified on an abdominal X-ray: clear delineation of both sides of the bowel wall by extraluminal air (Rigglers sign), linear gas streaks in the upper abdomen (falciform ligament sign), or loss of density overlying the liver (hepatic lucency), or odd-shaped air pockets may be evident on close inspection (Fig. 12.11).
If there is suspicion of urethral injury the integrity of the urethra can be easily checked in the resuscitation room by a retrograde urethrogram. The technique is described in Box 12.1. In the past, when renal injury was suspected, a one shot intravenous urogram (IVU) may have been performed, but this has largely been superseded by contrast enhanced CT.
When undertaken in the resuscitation room under local anaesthesia this is an imprecise technique with up to an 88% false positive rate. It is not recommended except when it can be performed in the operating theatre with the patient prepared and consented for proceeding to laparotomy as necessary.
Diagnostic peritoneal lavage (DPL)
DPL remained the gold standard investigation of blunt abdominal trauma for 30 years until the advent of readily-available CT and FAST scanning. Its role continues to be widely debated and whilst certainly no longer a first line investigation in most circumstances, it remains a useful, rapid, bedside investigation that can offer valuable information to the decision making process after abdominal injury, particularly if ultrasound and CT are unavailable.
The only definite contraindication to DPL is the presence of an indication for laparotomy, but DPL is relatively contraindicated in the uncooperative, the obese, children, pregnant women and those with previous multiple abdominal surgery. It may be performed by an open or Seldinger technique, which is quicker, but has a higher rate of complications. Insertion of both a nasogastric (or orogastric) tube and urinary catheter are mandatory before DPL. The main complications of DPL are gut injury, haemorrhage and intra-peritoneal infection. The technique is described in Box 12.2.
A macroscopically positive result is seen when:
• >10 mL of frank blood is drained; or
• Gastric contents are aspirated; or
• Urine is aspirated; or
• The lavage return contains bile or vegetable material.
If the result is not macroscopically positive then the results are determined by laboratory analysis. This is often a lengthy process and one of DPL’s greatest attributes (immediacy) is lost.
A microscopic positive result is seen when:
• RBC >100,000/mm3; or
• WCC >500/mm3; or
• Amylase >200 U/L.
Using these criteria, DPL after blunt trauma has a sensitivity and specificity of 94.4 and 99%, respectively, with an overall accuracy of 98.1%. When applied to penetrating trauma, these lavage counts yield an unacceptably low sensitivity, and the threshold for a positive result may have to be lowered to as low as RBC >1000/mm3 to achieve satisfactory results.
A positive DPL has traditionally mandated laparotomy and in circumstances where no further imaging is available this should still hold true. The non-therapeutic laparotomy rate after positive DPL was historically given as approximately 15%, but in an era where selective non-operative management of abdominal injury is increasingly commonplace, this would now undoubtedly be higher. Mandatory laparotomy after a positive DPL precludes the concept of non-operative management of abdominal injury and now represents one of the biggest drawbacks of the technique. In a stable patient, a positive DPL may be augmented by a CT scan to quantify the level of solid organ injury
Focused abdominal sonography in trauma (FAST)
Ultrasound is an excellent non-ionizing radiation modality for identifying solid organ injuries and intra-abdominal free fluid. However, its utility in the management of trauma patients is limited partly by the lack of high quality mobile machines, but more usually by the lack of availability of a trained sonographer. This is particularly noticeable outside normal working hours.
To overcome this difficulty, FAST has been developed as an easily applicable technique for non-radiologists using readily available portable equipment. It uses an ultrasound scanner, usually a 3.5–5.0 MHz convex transducer, in four defined positions on the abdomen to identify fluid in the hepatorenal and splenorenal pouches, and the pelvis (Figs 12.13 and 12.14). The fourth FAST position scans the lower mediastinum for evidence of pericardial effusion.
The technique is specifically designed to look for the presence of fluid in these four positions, and it is a rapid, easily repeatable bedside investigation that does not necessitate the patient being moved from the resuscitation room. It is not designed to identify solid organ injury. It is operator dependent and the view quality may be degraded by patient obesity and overlying bowel gas artefact. A small amount of pelvic free fluid in women of reproductive age is normal.
Studies have shown radiologists and non-radiologists who have been appropriately trained to be equally good at detecting free fluid. Volumes below 100 mL are unlikely to be identified by FAST, with only 10% of scanners able to detect 400 mL with a mean threshold volume for detection of 619 mL. For the assessment of blunt abdominal trauma in experienced hands FAST is reported to have sensitivity and specificity of 73–88% and 98–100%. In penetrating trauma, the evidence base is smaller, but the specificity remains extremely high, although the sensitivity falls to between 46–67%.
A positive FAST is a strong predictor of intra-abdominal injury, and when combined with haemodynamic instability necessitates laparotomy. In the stable patient, further imaging may be appropriate and consideration of conservative management. The problems arise when considering what to do with a negative FAST. There is currently insufficient evidence to safely allow discharge after a single negative FAST in abdominal trauma. If the patient is stable then they should undergo alternative imaging, ideally CT. An alternative is a repeated FAST scan—this may be performed after 30 min in the resuscitation room or after 6 h when admitted to a ward and this has been shown to increase the pick-up rate of occult injuries.
If abdominal assessment has not been completed in the resuscitation room then consideration can be made of continued investigation outside of the resuscitation room. This is usually in the form of computed tomography (CT) scanning. CT utility in managing both blunt and penetrating trauma has increased in line with the technological advantages in CT scanners, culminating in multi-slice CT scanning, which enables the entire abdomen to be scanned in a single breath hold, utilizing thinner slices, and reducing artefact.
CT scanning is indicated in abdominal trauma patients in whom suspicion of intra-abdominal injury remains after initial assessment, and who are haemodynamically stable. It is important to recognize the difference between haemodynamic stability and normality at this juncture. It is not necessary for a patient to exhibit normal vital signs to proceed to scanning, but they must have parameters that need minimal fluid infusion to maintain them. In addition, scanning of the abdomen should be considered when there are obvious injuries above and below the abdomen, a pattern which is highly suggestive of abdominal injury in between, and in the obtunded or unconscious patient in whom abdominal assessment is difficult or unreliable.
CT is the best modality for imaging the solid organs and retroperitoneum and gives useful information about bony injury, especially the pelvis where 60% of significant fractures are associated with visceral injury. It has to date been used predominantly in blunt trauma, as penetrating abdominal injury has traditionally been treated by mandatory laparotomy. With the advent of a more conservative approach in some centres, CT has been used to address the issue of peritoneal penetration, especially by tangential gunshot wounds (Fig. 12.15).
A non-contrast scan gives little information about the vascularity of organs or bleeding, but will identify free air within the peritoneum. Intravenous contrast and scanning in the portal venous phase will identify haematomas within solid organs and a delayed scan 3–5 min later will identify ongoing extravasation of blood. The addition of oral (or nasogastric) contrast may increase the detection rate of gastrointestinal injury, although the evidence is conflicting; it also risks aspiration and delaying the scan at least 30 min to allow contrast to pass through the small bowel is often not possible. It is perhaps of most use to delineate suspected duodenal injury, as contrast passes quickly and scans are not delayed. If there is a suspicion of lower gastrointestinal tract injury, especially the rectum and left colon, then rectal contrast may also prove useful (the so called triple-contrast scan).
In penetrating abdominal injury, the sensitivity and specificity of predicting the need for laparotomy are both over 96% in most studies and approach 100% in some. The greatest contribution of CT scanning to modern abdominal trauma management has been its ability to reliably define the degree of solid organ injury, enabling the concept of selective non-operative management (SNOM) of abdominal injury. The American Association for the Surgery of Trauma (AAST) has produced specific organ injury scales for all abdominal and thoracic organs, which in most instances depend on CT evaluation of the organ. The scale for liver injury is shown in Table 12.1.
Table 12.1. The AAST liver injury scaling system
Subcapsular <10% surface area
Capsular tear <1 cm parenchymal depth
Subcapsular 10–50% surface area; intraparenchymal <10 cm in diameter
Capsular tear 1-3 cm parenchymal depth, <10 cm in length
Subcapsular >50% surface area or ruptured subcapsular/parenchymal haematoma; intraparenchymal haematoma >10 cm or expanding
3 cm parenchymal depth
Parenchymal disruption involving 25–75% hepatic lobe or 1–3 Couinaud’s segments within a single lobe
Parenchymal disruption involving >75% of hepatic lobe or >3 Couinaud’s segments within a single lobe
Juxtahepatic venous injuries such as to retrohepatic vena cava/central major hepatic veins
Features of injury on CT scanning
Pneumoperitoneum If not evident on the trauma series chest radiograph, a CT scan will clearly demonstrate extraluminal air as a black shadow (Fig. 12.16). If doubt still exists, viewing the scan using lung windows will highlight extraluminal air more clearly.
Solid organ injury. Lacerations appear as linear hypodense areas, usually extending from one surface of the organ. Haematomas are often either centrally placed (Fig. 12.17A) or lie beneath the capsule (Fig. 12.17B) and are usually oval or round.
Bleeding. This may be evidenced by free fluid in the abdomen (blood has a CT density of 40 Hounsfield units) or by active leak of intravenous contrast which is best seen in the arterial phase of the scan.
Bladder injury. This is best demonstrated by a dedicated CT cystogram performed before intravenous contrast is given. 100 mL of dilute iodinated contrast is instilled via the urethral catheter and if no leakage of contrast is seen, further contrast is instilled. If a dedicated cystogram is not performed then bladder injury may be identified from the presence of intra-peritoneal free fluid, blood clots within the bladder or a deformation of the bladder shape.
Bowel injury. Identification of bowel injury is notoriously the weak point of CT scanning and many signs of bowel injury are indirect. The presence of free fluid without solid organ injury should raise the suspicion of gut perforation. Bowel wall thickening, mesenteric fat streaking and mesenteric haematoma are all predictors of bowel injury.
Magnetic resonance imaging
MRI scanning is rarely utilized in the acute management of abdominal trauma. It is often not available out of hours, and requires experienced MRI radiologists to interpret the images. The caveats that apply to transfer of a trauma patient to the CT scanner with regard to haemodynamic instability are equally if not more applicable to the acute use of MRI. Metal objects are not allowed within the scanner room and, thus, ventilation of a critically ill patient in the MRI suite requires specialist equipment not commonly available. MRI scanning may have a role to play in the investigation of the stable trauma victim in specific circumstances, such as pregnancy or renal failure where either high dose radiation or intravenous contrast media are contraindicated. MRI is also useful to delineate complex anatomy prior to reconstructive surgery, especially following hepatopancreaticobiliary injury.
Diagnostic laparoscopy in trauma is uncommon, but can be useful in certain circumstances. It is poor at visualizing the retroperitoneum and assessing the whole small bowel can be technically demanding and time-consuming. It is, however, excellent at assessing peritoneal penetration in stabbings or tangential gunshot wounds and at visualizing the diaphragms. If used to assess peritoneal violation any entry wounds should not be used as port sites due to the risk of restarting bleeding. Laparoscopy risks precipitating a tension pneumothorax or venous gas embolism if there has been significant vascular injury. Laparoscopy normally requires a general anaesthetic and should be performed in the operating theatre, although at least one study has used laparoscopy in the Emergency Department (ED) under local anaesthetic and sedation, to evaluate peritoneal breach after penetrating injury, facilitating discharge of 10 out of 16 patients from the ED. Therapeutic trauma laparoscopy is still in its infancy although laparoscopic splenectomy or splenic salvage surgery in isolated injury has been reported.
Trauma laparotomy is a demanding undertaking, and requires the input of experienced surgeons who must possess a wide range of surgical skills outwith those in routine practice. The basic technique is described in Box 12.3.
Specific details of the many techniques required to deal definitively with many of the injuries mentioned are beyond the scope of this chapter (see Chapter 19 and further reading). The need to convert to a damage control approach should always be borne in mind when embarking on a trauma laparotomy.
Damage control surgery
Mortality and morbidity from definitive surgery in the abdominal trauma victim has historically been high. The early surgical literature contained passing references to a concept of not operating in some types of severe injury, and packing of hepatic injuries (with a reasonable degree of success) was reported by Pringle in 1908 and Halsted in 1913, but fell from favour during the Second World War, due to further bleeding on pack removal and problems with sepsis attributed to the packs. Thereafter, sporadic successes from hepatic packing were reported, but it was the work of Rotondo and colleagues in 1993 that most clearly demonstrated the potential advantage of a damage control approach in the most critically ill trauma victims. Although he reported no overall significant difference between the two groups who had definitive or damage control laparotomy, when the subset of the most severely injured was analysed, DC had a significantly improved survival rate (10/13 = 77%) compared with (1/9 = 11%; p = 0.02) with traditional definitive laparotomy.
The basis of damage control surgery (DCS) is that it is the deranged physiology of the trauma victim that is responsible for the poor outcome and this is exacerbated by stress of embarking on lengthy definitive surgery.
Pathophysiology of DCS
• Coagulopathy: the coagulopathy of trauma has a dual aetiology. Exsanguinating haemorrhage depletes the reserve of clotting factors, which is further diluted by large volume fluid resuscitation. This, in turn, causes platelet and coagulation factor dysfunction and activation of the fibrinolytic system leading to a hypocoagulable state. Ongoing blood loss diminishes cellular perfusion and contributes to acidosis and hypothermia.
• Acidosis: prolonged haemorrhagic shock leads to cellular hypoperfusion, anaerobic cellular metabolism, and lactic acid production. This produces a profound metabolic acidosis, which reduces the efficiency of the clotting cascade, and promotes coagulopathy and blood loss.
• Hypothermia: hypothermia is a consequence of severe exsanguinating injury and subsequent resuscitative efforts, as well as climatic effects before arrival in hospital. Severe haemorrhage leads to decreased tissue perfusion and diminished oxygen delivery, a result of which is reduced heat generation.
These three physiological derangements are synergistic and have been termed the ‘bloody vicious triad’ of trauma (Fig. 12.18).
In short, the concept of DCS begins with an abbreviated laparotomy aimed at minimizing the metabolic insult, rather than restoration of anatomic integrity by temporary cessation of haemorrhage and limitation of gastrointestinal contamination, followed by admission to the Intensive Care Unit for restitution of physiology followed 24–72 h later by a return to theatre and completion of definitive surgical repair of the injuries.
Indications for damage control laparotomy
The decision for DCS should ideally be taken before laparotomy is started, by recognition of the poor physiological state of the patient during the resuscitation phase, when the need for laparotomy is first identified, or identification of injury complexes that would adversely affect the patient’s physiology if definitive treatment were undertaken.
Surgical triggers for DCS are:
• Inability to achieve haemostasis.
• Combined vascular, hollow, and solid organ injury.
• Inaccessible vascular injury.
• Lengthy operative procedures.
• Need for surgical treatment of other injuries, e.g. craniotomy for intra-cranial haematoma.
• Need for non-surgical control of other injuries, e.g. angio-embolization of pelvic haemorrhage.
• Inability to close the abdomen or surgical reason for re-look laparotomy.
Physiological triggers for DCS are:
• pH < 7.2.
• Core temperature < 34°C.
• Coagulopathy (prothrombin time >16 s or activated partial thromboplastin time >60 s).
• Serum lactate >5 mmol/L.
• Probable operation time >60 min.
• 10 unit blood transfusion.
• Systolic BP < 90 mmHg for more than 1 h.
Technique of damage control laparotomy
The patient should be fully prepared and draped from the nipples to the knees to allow exposure and extension into the junctional areas and chest if necessary. The three main principles of the laparotomy are to arrest haemorrhage, limit contamination and temporarily close the abdomen.
Arrest bleeding: intra-abdominal free blood and clots are rapidly evacuated and four quadrant packing performed, or if haemorrhage is torrential a supra-hepatic aortic clamp may be applied. Packs are then removed, starting in the quadrant thought least likely to harbour significant bleeding. Haemorrhage is temporarily controlled when found as definitive treatment may waste time on what may not be the major source of haemorrhage.
Liver bleeding can be reduced by bimanual compression or Pringles manoeuvre. These are unlikely to stop liver bleeding definitively and the commonest DCS manoeuvre is firm liver packing after mobilization of the liver by division of its supporting ligaments. If a single penetrating liver track is bleeding balloon tamponade by a Foley catheter can be used (Fig. 12.19).
If the spleen is bleeding a swift splenectomy is appropriate rather than wasting time on temporary haemorrhage control. Retroperitoneal haemorrhage not controlled by its own tamponade requires visceral mobilization for access (Cattell–Braasch or Mattox manoeuvres). Major vessel injury can be temporarily shunted.
Limit contamination: gastrointestinal contamination is limited by temporary exclusion of damaged segments of bowel. This can be rapidly achieved by soft bowel clamps, nylon tapes or linear stapler cutters. Biliary and pancreatic leakage can be managed emergently by drainage in the first instance either by sump drains or a biliary T-tube if easily placed.
Temporary abdominal closure: temporary abdominal closure (TAC) is a sensible technique because DCS mandates re-look laparotomy, and definitive closure is time consuming and markedly increases the risk of abdominal compartment syndrome (ACS). It may be achieved by suture of the skin alone or more commonly by application of a Bogota Bag (Fig. 12.20a, b) or Opsite sandwich closure.
Abdominal compartment syndrome
An acute rise in intra-abdominal pressure leads to a series of well defined physiological consequences, which have been termed the abdominal compartment syndrome (ACS). The aggressive fluid resuscitation of trauma is a well-recognized contributing factor to the development of ACS. Increased intra-abdominal pressure reduces venous return and decreases cardiac output; decreased perfusion pressure and direct compression of abdominal vessels reduces splanchnic and renal perfusion, which manifests as gut acidosis and oliguria. Splinting of the diaphragms increases intra-thoracic pressure and requires increased ventilatory pressures to maintain oxygenation.
Treatment is by reduction in intra-abdominal pressure; whilst some medical measures, such as decompression of the intestinal contents by enemas or nasogastric aspiration, complete muscle relaxation, and the use of osmotic agents to offload abdominal oedema may help, the mainstay of management is surgical decompression of the abdomen. This may then be left open as a laparostomy (with or without temporary coverage from a Bogota bag or Opsite dressing) or closed temporarily with a mesh sutured to the abdominal fascia. It should be noted that ACS may recur even after TAC.
Resuscitation in intensive care
On completion of the damage control laparotomy the patient is transferred to ICU for restitution of normal physiology. Rewarming by electric blankets and warm intra-venous fluids can be augmented by warmed bladder or thoraco-abdominal cavity irrigation in severe hypothermia. Optimization of cellular perfusion is achieved by restoration of circulating volume with crystalloid and blood products coupled with supplemental oxygenation and inotropes as required. Aggressive replacement of clotting factors by infusion of platelets, fresh frozen plasma and cryoprecipitate is used to correct coagulopathy. The timing of re-laparotomy depends upon the speed with which normal physiological parameters can be restored, but is normally scheduled for 24–48 h after initial surgery. If there is evidence of ongoing haemorrhage, contamination or development of ACS (which may occur even in the presence of TAC) then the return to theatre may be brought forward for further damage control.
Second look laparotomy
When the patient is normothermic with neither coagulopathy nor acidosis then they can be returned to theatre for definitive repair of injury. Packs are removed and any re-bleeding attended to; vascular shunts are replaced by definitive interposition grafts. Gastrointestinal injury can be resected and continuity restored or stoma created. Hepatopancreaticobiliary injury may be best served by surgery from a specialist as reconstructive surgery can be complex—this may necessitate transfer to a specialist centre prior to re-look laparotomy. In the absence of specialist skills T-tube drainage of the biliary tree is safe, as is the placement of large sump drains adjacent to an injured pancreas. In the face of pancreatic duct injury, definitive control of the duct is required otherwise a persistent systemic inflammatory response ensues, adding significantly to the mortality and morbidity. A complete inspection of the abdomen is repeated to detect missed injuries. It is often impossible to definitively close the abdomen at this stage and a variety of temporary abdominal closure techniques including vacuum systems, meshes and Velcro patches are available.
Several decades ago haemodynamic instability after blunt abdominal trauma or any penetrating mechanism of abdominal injury was a mandate for laparotomy. Historically, this has led to a negative and non-therapeutic laparotomy rate of approximately 25%. It was thus realized that not all abdominal injuries injure the contents of the abdomen and that not all injuries within the abdomen require surgical intervention. This revelation coupled with the high complication rate (up to 41% following non-therapeutic trauma laparotomy) and attendant financial costs, are the underlying rationale for the advent of selective non-operative management (SNOM). It was initially applied to blunt injuries only, and it is only recently that non-operative management of stab wounds to the abdomen has been adopted. The incidence of intra-abdominal injury after gunshot wounding is about 98% in most studies, and so non-operative management of gunshot wounds remains a controversial technique although high volume trauma centres such as the LA County trauma room report successful non-operative management in up to 30% of abdominal gunshot injuries.
Selective non-operative management of abdominal injuries now has fairly standard requirements as patients managed in this way have an enormous capacity for rapid, catastrophic deterioration. The following general criteria should be met before consideration of SNOM:
• Appropriate injuries (Grade I–III solid organ injuries on CT).
• No suspicion of hollow viscus injury.
• Cardiovascular stability (and an acute transfusion requirement of <2 units).
• Minimal physical signs.
• Availability of high dependency or critical care facilities for observation.
• Patient available for repeated frequent reassessment (preferably by the same senior clinician) to detect subtle changes in condition. Thus, patients requiring urgent surgery for non-abdominal injuries such as fracture fixation will not be available for repeated evaluation and must be excluded from non-operative management strategies.
Individual organ injuries generate their own criteria for SNOM and should be considered, as well as these general criteria. The investigation that underpins selection for SNOM is CT scanning, and thus by definition patients must be stable enough to undergo CT; this automatically precludes the haemodynamically unstable, which are usually the high grade injuries. Certain CT features, such as contrast ‘blush’ or active extravasation suggest that SNOM without intervention is unlikely to succeed.
SNOM may be achieved without any intervention, but the advent of more readily available interventional radiology has revolutionized non-operative management.
Basic technique of interventional angiography in trauma
Angiography requires a degree of haemodynamic stability as ongoing resuscitation is difficult in a radiographic screening room. Critical care facilities should be available in the angiography suite. Arterial access is usually gained by the femoral or brachial routes.
If active extravasation is demonstrated then embolization can be achieved by instillation of coils, gelfoam particles or a combination of both. The more selective (and, therefore, distal) the embolization, the smaller the territory of ischaemia that is induced. If superselective embolization is not possible, for instance in the pelvis, then more proximal control of the internal iliac arteries may be indicated to obtain haemodynamic control, accepting the increased risks of collateral ischaemic injury.
Significant bony pelvic injury is associated with a high risk of massive bleeding, usually contained within the loose tissues of the retroperitoneum, and this may be arterial or venous. Interventional radiology is now the mainstay of treatment. If CT scanning shows evidence of ongoing pelvic bleeding, such as contrast extravasation, bladder compression from haematoma or continuing transfusion requirement in the absence of other causes, then therapeutic angiography is required. If free intra-abdominal bleeding occurs, or temporizing measures, such as fracture stabilization by pelvic wrapping or external fixation fail to achieve haemodynamic stability then the patient may require immediate DCS with extraperitoneal pelvic packing to establish stability before diagnostic angiography +/– arterial embolization.
Indications for angiography in blunt splenic injury include contrast extravasation (extrasplenic) or blush (intrasplenic), falling haemoglobin with known splenic injury and pseudoaneurysm formation. Superselective embolization of the distal vessels limits the area of ischaemic injury and maximizes the volume of functional splenic tissue. In cases where superselective embolization is not possible embolization of the splenic artery will significantly reduce the arterial pressure within the spleen, which may be sufficient to allow spontaneous arrest of haemorrhage, whilst maintaining perfusion from collateral vessels. Such a reduction of the perfusion pressure may also allow an operative splenic preservation technique to be achieved if embolization is insufficient and the patient ultimately requires surgery.
The increasing success of SNOM for splenic injury is reflected in the literature. A large retrospective review of nearly 1500 patients with blunt splenic injury identified attempted SNOM in 55% of cases with an 11% failure rate; increasing grade of splenic injury, increasing degree of haemoperitoneum and an Injury Severity Score >15 were all identified as predictors of failure of SNOM in blunt splenic injury. Interestingly, of those who failed SNOM, two-thirds did so within the first 24 h and only 10% went on to have splenic conservation surgery. Many reports suggest that the incidence of splenic salvage after SNOM is as high as 90% but these reports, as the EAST report does, count only those considered for SNOM—overall the rate of organ preservation is between 45 and 60%. The overall rate is much higher in children (∼85–90%), potentially because the splenic capsule is relatively thicker in children, tears tend to lie parallel to the distribution of the blood vessels, thereby limiting haemorrhage, and children have a greater physiological reserve and no pre-morbid conditions to limit SNOM.
The indications for angiography in blunt hepatic trauma are similar to those for splenic injury and the same caveats apply. Most blunt liver injuries are low pressure venous injuries and will stop bleeding spontaneously, but contrast blush on CT suggests hepatic arterial injury and is likely be suitable for embolization using techniques similar to those for the spleen, aiming to place coils as peripherally as possible. A high grade hepatic injury is not a contraindication to potential SNOM, but it must be realized that the chances of failure are greater. Low grade injuries have a failure rate of 3–7.5% compared with 14% for grade IV and 22.6% for grade V injuries. Arterial embolization may be employed after liver packing at DCS.
Complications of SNOM of liver injury
• Bile leaks: occur in up to 20% of liver injuries, but can be managed without recourse to surgery. The majority can be treated by a combination of percutaneous drainage and endoscopic management of the biliary ducts.
• Hepatic abscess: these are a relatively rare complication and, again, are usually managed by percutaneous drainage. Persistent or recurrent infected collections will require laparotomy.
• Delayed haemorrhage: the advent of interventional radiology early in the management pathway has reduced the previously high rate of delayed bleeding from rupture of subcapsular haematoma; such delayed bleeds can be equally well treated by delayed embolization.
• Ischaemia: may occur as a result of the injury itself or embolization may disrupt vascular supply to hepatic segments giving rise to rising liver enzyme levels, abdominal pain and sepsis—resection of the dead segments is indicated.
Abdominal SNOM and head injury
Attempted non-operative management is not contraindicated by a concomitant head injury. The presence of neurotrauma increases the likelihood of disseminated intravascular coagulopathy which may exacerbate any intra-abdominal bleeding, so coagulopathy should be aggressively treated in these circumstances. Head injury increases the likelihood of SNOM failure but should not deter the attempt.
Surgical decision-making in abdominal trauma can be complex, and should be performed by the most senior and experienced surgeon available. It is also true to say that decisions are not final and should be re-evaluated in light of each new piece of information or change in the patient’s clinical condition.
Choice and timing of investigation
Which investigation to choose, and when, is largely based on individual operator experience, local availability of resources and, most importantly, the haemodynamic stability of the patient. The following rules are widely applicable and should encompass all clinical scenarios.
• Catastrophic haemodynamic collapse with abdominal signs: the patient is moribund and delay may prove fatal. The primary survey should be cut short for immediate resuscitative laparotomy. If the abdomen is negative then thoracotomy should be considered with cross-clamping of the descending aorta to achieve haemodynamic stability.
• Gross haemodynamic instability with abdominal signs: investigation is contraindicated and immediate laparotomy is required.
• Gross haemodynamic instability without abdominal signs: if a likely alternative source of haemorrhage is clear, such as obvious thoracic injury or multiple long bone fractures, then rapid resuscitation room abdominal assessment by FAST (or DPL) is appropriate. If this is negative, treatment priorities lie elsewhere. If positive, or no other likely cause of the ongoing haemodynamic compromise is evident then laparotomy should follow.
• Haemodynamic instability with temporary response to fluid resuscitation (transient fluid responders): if they are too unstable to move from the resuscitation room to the CT scanner then FAST (or DPL) is required. If positive and the patient remains unstable then laparotomy is indicated. If haemodynamic stability is achieved after a positive test then further imaging by CT will allow accurate grading of solid organ injury. If FAST is negative and haemodynamic stability is achieved, then the patient should be admitted and observed. In the face of haemodynamic compromise and a negative test then the decision lies between further imaging such as CT if stability allows, a period of further resuscitation with close monitoring and repeat test, or if haemodynamic parameters are worsening then laparotomy is permissible despite normal tests.
• Stable patient with signs of abdominal injury OR an equivocal abdominal examination in the face of minor injuries: these patients should undergo CT evaluation and the decision on observation, SNOM or laparotomy made on the basis of the findings.
Who needs laparotomy?
The indications for laparotomy have changed over the years and now depend on the degree of experience of the managing clinicians with abdominal trauma, as some experienced trauma physicians will now manage many cases of penetrating trauma conservatively.
• All abdominal gunshot wounds (excluding those where the patient is stable and CT convincingly demonstrates an extraperitoneal wound track).
• Unstable patients after resuscitation with abdominal signs of trauma
• Unstable patients after resuscitation with a positive FAST (or DPL).
• Positive CT findings unsuitable for conservative management.
• Intra-abdominal haemorrhage uncontrolled by interventional radiology.
• Abdominal stab wounds with evidence of peritoneal penetration.
Thoracotomy or laparotomy first?
In cases where there is actual or potential thoraco-abdominal injury the decision must be made as to which cavity is attended to first. A large study from USA has indicated that penetrating thoraco-abdominal trauma is associated with a high mortality (31%) and this is almost doubled (59%) if there is a need to explore both cavities. The authors assessed how many cavity procedures had to be interrupted to access another cavity (an indication of incorrect sequencing), which occurred in 36% of cases where thoracotomy was the first procedure compared with 53% of laparotomy first procedures. The commonest reason for inappropriate initial thoracotomy was a high thoracostomy output derived from abdominal injuries draining via a diaphragmatic rent.
Duration of observation period?
Whilst the indications and techniques for SNOM of solid organ injury are now well established, what is less clear is how long patients should be observed in hospital, and for how long they should be advised to refrain from strenuous activity, especially contact sports. CT evidence of organ healing appears to be related to the grade of injury with the more severe grades of injury taking longer to heal. The recommendations of the American Paediatric Surgical Association are that grades I, II, III, and IV liver and spleen injuries should be observed in hospital for 2, 3, 4, and 5 days, and refrain from strenuous activity for 3, 4, 5, and 6 weeks, respectively. No such guidance exists for adult patients. Routine re-imaging is not indicated for either adult or paediatric patients managed conservatively after blunt trauma; all imaging is directed by changes in haemodynamic status.
Haemodynamic instability with a pelvic fracture
As pelvic fractures have the potential for both intra- and retro-peritoneal haemorrhage it is important to ascertain the likely cause of instability urgently. Vertical and lateral compression injuries are usually associated with a lower degree of blood loss, which is commonly extraperitoneal. Instability in the face of these fracture patterns should prompt a search for other injuries. In both stable and unstable pelvic fracture patients pelvic stabilization should be promptly instigated; a pelvic binder or sheet is probably the easiest and quickest technique if the fracture pattern is suitable for this approach. In stable patients CT is the investigation of choice; in the unstable, a resuscitation room evaluation of the abdomen is necessary by FAST (or DPL). Peritoneal lavage in the face of a pelvic fracture should be performed by an open technique and in a supra-umbilical site to reduce the incidence of false positive tests—a negative DPL rules out the abdominal cavity as the source of haemodynamic instability. A positive FAST or DPL, plus instability merits laparotomy and if there is a large pelvic haematoma, extraperitoneal packing is used to achieve stability. At whatever stage haemodynamic stability is achieved, or the FAST is negative, pelvic angiography should be used to identify and embolize the source of haemorrhage. When definitive haemorrhage control has been achieved and physiology normalized, attention is then turned to more formal stabilization of the pelvic fracture.
Combined abdominal and head trauma
This combination of injuries is relatively uncommon, with large series suggesting an incidence of 5.7–13.4% of severe blunt head injury (GCS<8) and abdominal injuries on CT scanning. When present in combination, there are difficulties in both assessment and management. Neurological impairment will reduce the sensitivity of abdominal examination, and the permissive hypotension often used in the non-operative management of solid organ abdominal injury will aggravate secondary brain injury.
In the unstable patient, laparotomy takes precedence irrespective of neurological status, but consideration must be given to intra-operative intracranial pressure monitoring or burr holes if the GCS <9 or there were lateralizing signs before anaesthesia. Injuries to the abdomen and head are unlikely to have spared the intervening torso and the assessing clinician should be aware of the possibilities of thoracic injury as well.
American College of Surgeon Committee on Trauma. Advanced Trauma Life Support (ATLS®) Chicago: ACS, 2007. Available at: www.trauma.org
Amin SN, Rowlands BJ. Colorectal trauma. Trauma 2000; 2: 211–21.Find this resource:
Boffard KD (ed.). Manual of Definitive Surgical Trauma Care, 2nd edn. London: Hodder Arnold, 2007.Find this resource:
Bowley DMG, Barker P, Boffard KD. Damage control—concepts and practice. J Roy Army Med Corps 2000; 146: 172–82.Find this resource:
Feliciano DV, Mattox KL, Moore EE. Trauma, 6th edn. New York: McGraw-Hill, 2008.Find this resource:
Herr BW, Gagliano RA. Historical perspective and current management of colonic and intraperitoneal rectal trauma. Curr Surg 2005; 62(2): 187–92.Find this resource:
Nahum AM, Melvin J. Accidental Injury: Biomechanics and Prevention. Berlin: Springer, 2001.Find this resource:
Pietzman AB, Heil B, Rivera, et al. Blunt splenic injury in adults: multi-institutional study of the Eastern Association for the Surgery of Trauma. J Trauma 2000; 49: 177–89.Find this resource:
Rotondo M, Schwab CW, McGonigal MD, et al. Damage control: an approach for improved survival in exsanguinationing penetrating abdominal injury. J Trauma 1993; 35(3): 375–82.Find this resource:
Paterson Brown S. A Companion to Specialist Surgical Practice: Core Topics in General and Emergency Surgery, 3rd edn. London: Elsevier Sanders, 2008.Find this resource:
Stylianos S, and the APSA Trauma Committee. Evidence based guidelines for resource utilization in children with isolated spleen or liver injury. J Paediat Surg 2000; 35: 164–9.Find this resource:
Whitfield C, Garner JP. The early management of gunshot wounds part II: the abdomen, extremities and special situations. Trauma 2007; 9: 47–71.Find this resource:
Whitfield C, Garner JP. The early management of gunshot wounds part II: the abdomen, extremities and special situations. Trauma 2007; 9: 47–71.Find this resource:
Whitfield CG, Garner JP. Beyond splenectomy—options for the management of splenic trauma. Trauma 2008; 10: 247–59.Find this resource: