Surgery is the primary treatment modality for managing intra-abdominal infection caused by perforation of a hollow viscus, transmural necrosis of the gastrointestinal tract, or pancreatic necrosis. The elimination or control of the infectious source and the reduction of peritoneal contamination by debridement and lavage are the cornerstones of primary operative treatment.
In general, the source of intra-abdominal sepsis is eliminated by resection, closing the perforated viscus or resection of any necrosis. For pathology located in the large intestine, resection is usually performed followed by creation of a proximal enterostomy (Hartmann's procedure). In this location, primary anastomosis is avoided because of the high risk of anastomotic dehiscence. The risk associated with primary anastomosis of the small intestine following resection of a diseased segment is considered to be much lower. However, if peritoneal soiling is particularly extensive, resection plus proximal and distal enterostomy is advocated. Suturing a perforation is now mainly performed for perforated peptic ulcer. Additionally, it may be considered for a single small foreign body or traumatic perforation of the small intestine, if diagnosed early.
Reduction of peritoneal contamination by debridement and intra-operative lavage is intended to prevent residual infection. As discussed, adjuvant substances such as foreign material, necrotic tissue, fibrin, bile, blood, or intestinal contents enhance the severity of the infectious process by stimulation of bacterial growth and impairment of neutrophil and macrophage function. Radical debridement—eliminating all adherent fibrinous deposits on peritoneal surfaces—was initially promising but later criticized because of the high risk of excessive bleeding from serosal surfaces and intestinal perforations. Intra-operative lavage has become a standard procedure, although the efficacy of intra-operative lavage has not been well documented in clinical studies. The addition of antibiotics or antiseptics in the lavage solution does not improve outcome in patients with intra-abdominal sepsis who are already receiving appropriate systemic antibiotics.
There is a great variance in surgical strategies to be followed after elimination of the source, debridement, and intra-operative lavage, particularly in patients with severe intra-abdominal sepsis. These strategies vary from a conservative 'wait and see' policy to the most aggressive method of 'planned relaparotomy'. Four strategies can be distinguished: a 'wait and see' policy with or without tube drainage, continuous postoperative peritoneal lavage, 'open' drainage (laparostomy), and 'planned relaparotomy'.
In the 'wait and see' policy, surgeons rely on host defenses to eradicate bacteria and adjuvant substances that remain intraperitoneally after the first operation. However, in severe forms of intra-abdominal infection, host defenses are unlikely to do this. This is illustrated by the high incidence of residual intra-abdominal infection in patients with diffuse peritonitis treated according to the 'wait and see' policy ( McLauchlan etal 1995). In this policy, surgeons also rely on clinical signs of residual or recurrent intra-abdominal infection in deciding whether or not to re-explore the abdominal cavity (i.e. 'on demand' relaparotomy). Clinical signs of residual intra-abdominal infection are often blunted in patients who are critically ill and cared for in intensive care units. Furthermore, diagnostic investigations such as ultrasound or CT for the detection of residual infection are often inconclusive or impossible to perform. Therefore, most 'on demand' relaparotomies are performed on the basis of otherwise unexplained progressive organ dysfunction or bacteremia. These relaparotomies are technically difficult, have a high morbidity, and often do not reverse organ dysfunction even when infectious foci are encountered and drained.
The role of tube drainage appears to be minor in diffuse intra-abdominal sepsis, mainly because of the rapidity with which drains are sealed off by fibrin deposition. Tubes placed intraperitoneally may cause visceral and vessel wall erosion with fistula formation and bleeding. Moreover, drains impair neutrophil function and potentiate infection. The only indications for drain insertion are evacuation of well-defined abscesses and provision of a preferential pathway for escape of visceral secretions such as bowel contents, pancreatic juice, and bile.
The concept of continuous postoperative peritoneal lavage has recently received more attention as a means of reducing residual intra-abdominal infection. At the time of initial surgery several peritoneal tubes are placed in 'strategic' positions to irrigate the abdominal cavity in the postoperative phase. Lavage with 10 to 20 liters of dialysis solution is advised to ensure dispersion of the solution throughout the abdominal cavity and to prevent fluid loculi. The mechanical action of continuous postoperative peritoneal lavage detaches necrosis, debris, and fibrin from surrounding tissues. Despite the potentially beneficial effects of continuous postoperative peritoneal lavage, reports on this method show conflicting results. Only for necrotizing pancreatitis has continuous postoperative peritoneal lavage been found a valuable adjunct to debridement of pancreatic necrosis.
The third strategy, leaving the abdomen open (laparostomy), was initially applied in France. The laparostomy should allow free drainage of exudate and simple access to the abdominal cavity after the first operation. In practice, free drainage becomes impossible within 1 to 2 days as a consequence of rapid adhesion formation by fibrin deposition. For the same reason, safe access to all parts of the abdomen will be difficult. The main advantage of this technique seems to be abdominal decompression. Leaving the abdomen open carries a risk of evisceration of abdominal contents, fistula formation, and abdominal wall hernia. In conjunction with nursing problems, these are considered the major limitations of this technique. These drawbacks have generated the application of alternative methods combining abdominal decompression with covering of intra-abdominal organs by, for example, Velcro analog (Sigma Medical, Apeldoorn, The Netherlands).
The concept of 'planned relaparotomy' has evolved from the technique of leaving the abdomen open. Re-exploration of the abdominal cavity is performed every 24 to 48 h to remove residual infectious material, such as necrosis, fibrin, and bacteria, and to prevent fluid loculi. The abdomen may be left open or temporarily closed by means of a zip or Velcro analog to relieve intra-abdominal pressure and facilitate re-exploration ( Fig 4).
Fig. 4 The Velcro analog (artificial burr), consisting of two sheets of polypropylene-nylon material, sutured to the fascia in a patient with severe intra-abdominal sepsis.
The success of this approach has been attributed to improved elimination of the bacterial inoculum, necrotic material, and fibrin, and the possibility of early detection and correction of complications. Favorable results compared with more conventional surgical therapies in terms of mortality and residual abscess formation have been reported, although prospective randomized studies are still lacking. More recently, disadvantages have been brought to light: unnecessary re-explorations, bleeding due to frequent manipulation of abdominal contents and removal of fibrin from serosal surfaces, 'spontaneous' fistula formation, interference with anastomotic healing by detaching adhesions and removing fibrin deposits, excessive fluid and protein losses, and large abdominal wall defects.
The ideal operative approach for severe intra-abdominal sepsis has yet to be established. More conventional approaches, such as 'wait and see' or continuous postoperative peritoneal lavage, do not seem to prevent residual or recurrent intra-abdominal infection and are associated with a high mortality. The method of 'planned relaparotomy' seems to decrease the rate of residual intra-abdominal infection. However, this method has a high complication rate and improvement in patient survival has yet to be clearly shown.
In some critically ill patients with the multiple organ dysfunction syndrome (MODS), signs of intra-abdominal sepsis persist in the absence of well-defined foci of residual infection or after successful surgical treatment of intra-abdominal infection. It has been hypothesized that bacteria in the gastrointestinal tract represent an undrained abscess that leads to the development of MODS. Interaction between these bacteria and cells of the monocyte-macrophage lineage releases cytokines and other cellular products which act as the effectors of end-organ dysfunction. The association between the disease processes leading to MODS and the egress of bacteria from the gastrointestinal tract (bacterial translocation) has been well established in animal studies ( Wells e.L§L 1988). However, clinical studies do not unequivocally support the relation between bacterial translocation and MODS. It is obvious that administration of antibiotics or a reoperation without clear indication will not alter outcome in these patients. Other approaches such as reducing bacterial translocation, modulation of cytokine release, and alteration of the target-organ effects of cytokines are the subject of study at present. Mortality will depend on the ability to restore immunological host defense mechanisms.
Was this article helpful?