Initial operation
The timing and adequacy of source control are of outmost importance in the management of intra-abdominal sepsis, as late and/or incomplete procedures may have severely adverse consequences on outcome.
Source control encompasses all measures undertaken to eliminate the source of infection, reduce the bacterial inoculum and correct or control anatomic derangements to restore normal physiologic function [
82,
83].
This generally involves drainage of abscesses or infected fluid collections, debridement of necrotic or infected tissues and definitive control of the source of contamination.
It is well known that inadequate source control at the time of the initial operation has been associated with increased mortality in patients with severe intra-abdominal infections [
84].
Early control of the septic source can be achieved using both operative and non-operative techniques.
An operative intervention remains the most viable therapeutic strategy for managing intra-abdominal sepsis in critical ill patients.
The initial aim of the surgical treatment of peritonitis is the elimination of bacterial contamination and inflammatory substances and prevention or reduction, if possible, of fibrin formation.
Generally, the surgical source control employed depends on the anatomical source of infection, the degree of peritoneal inflammation and generalized septic response, and the patient’s pre-morbid condition.
Surgical source control entails resection or suture of a diseased or perforated viscus (e.g. diverticular perforation, gastroduodenal perforation), removal of the infected organ (e.g. appendix, gallbladder), debridement of necrotic tissue, resection of ischemic bowel and repair/resection of traumatic perforations with primary anastomosis or exteriorization of the bowel.
Laparotomies are usually performed using a midline incision.
The primary objectives of surgical intervention include a) determining the cause of peritonitis, b) draining fluid collections, c) controlling the origin of the abdominal sepsis.
Special attention should be given to areas where abscesses may form such as the pelvis, the para-colic gutters, and the subphrenic spaces. These areas should be carefully exposed and debrided, avoiding bleeding by excessive peeling of the fibrin, and drained.
In case of suspected gastro-intestinal perforation, the whole extent of the GI tract, starting from the gastroesophgeal junction to the lower rectum should be thoroughly and carefully examined. If no perforation is found, the gastrocolic omentum should always be opened to expose the lesser sac to allow visualization of the posterior wall of stomach for any hidden perforation as well as careful examination of the body and tail of pancreas.
Special attention should be paid while draining and debriding the left subphrenic space since there is high risk of splenic injury during surgical manipulation due to fibrinous adhesions with the splenic capsule. Splenic bleeding maybe difficult to control due to adhesions and might warrant splenectomy which adds to the morbidity and potential mortality in an already compromised patient.
Intra-abdominal lavage is a matter of ongoing controversy. Some authors have favoured peritoneal lavage because it helps in removal as well as in dilution of peritoneal contamination by irrigation with great volumes of saline [
85]. However, its application with or without antibiotics in abdominal sepsis is largely unsubstantiated in the literature [
86].
In recent years, laparoscopy has been gaining wider acceptance in the diagnosis and treatment of intra-abdominal infections.
Laparoscopic approach in the treatment of peritonitis is feasible and effective without any specific complications in experienced hands. Laparoscopy has the advantage to allow, at the same time, an adequate diagnosis and appropriate treatment with the less invasive abdominal approach [
87]. However, in unstable patients laparoscopy is generally avoided because increased intra-abdominal pressure due to pneumoperitoneum seems to have a negative effect in critical ill patients leading to acid–base balance disturbances, as well as changes in cardiovascular and pulmonary physiology [
88].
Relaparotomy strategy
In certain circumstances, infection not completely controlled may trigger an excessive immune response and sepsis may progressively evolve into severe sepsis, septic shock, and organ failure [
89].
Such patients would benefit from immediate and aggressive surgical treatment with subsequent re-laparotomy strategies, to curb the spread of organ dysfunctions caused by ongoing sepsis.
Unfortunately, early assessment of the severity of peritonitis is difficult in emergency surgical patients and none of the existing and widely used ‘severity-of-disease’ scores, specifically developed for critically ill patients, were clinically useful in the identification of patients with ongoing infection needing a re-laparotomy [
90]. Surgical strategies following an initial emergency laparotomy include subsequent “re-laparotomy on demand” (when required by the patient’s clinical condition) as well as planned re-laparotomy in the 36-48-hour post-operative period.
On-demand laparotomy should be performed only when absolutely necessary and only for those patients who would clearly benefit from additional surgery.
Several studies have evaluated clinical variables that may be associated with the need for on-demand re-laparotomy in the immediate post-operative period [
91‐
97].
Van Ruler et al. [
92] in 2008 reported the results of a questionnaire asking surgeons to rank the importance of 21 clinical variables on their decision to re-operate in patients with secondary peritonitis. They found that diffuse extent of the abdominal contamination, localization of the infectious focus (upper gastrointestinal tract including small bowel), and both, extremely low and high leukocyte counts, independently predicted a re-laparotomy. These variables had only moderate predictive accuracy. The results of the questionnaire demonstrated that there was no consensus among surgeons about which variables are important in the decision-making process for re-laparotomy. The final decision to perform a re-operation on a patient in the on-demand setting is generally based on the patients generalized septic response and on the lack of clinical improvement.
Performing a case–control study, Koperna and Schulz [
91] retrospectively reviewed 523 consecutive patients with secondary peritonitis. They focused their attention on 105 patients, in whom standard surgical treatment of secondary peritonitis failed and who had to undergo re-laparotomy for persisting abdominal sepsis (study group). The authors showed that patients re-operated on after 48 hours had a significantly higher mortality rate than those operated on earlier (76.5% versus 28%; p < .001).
Planned relaparotomies, on the other hand, are performed every 36–48 hours for purposes of inspection, drainage, and peritoneal lavage of the abdominal cavity.
The concept of a planned relaparotomy for severe peritonitis has been debated for over thirty years. Re-operations are performed every 48 hours for reassessing the peritoneal inflammary process until the abdomen is free of ongoing peritonitis; then the abdomen is closed. The advantages of the planned re-laparotomy approach are optimization of resource utilization and reduction of the potential risk for gastrointestinal fistulas and delayed hernias.
The results of a clinical trial published in 2007 investigating the differences between on-demand and planned re-laparotomy strategies in patients with severe peritonitis found few advantages for the planned re-laparotomy strategy; however, the study mentioned that this later group exhibited a reduced need for additional re-laparotomies, decreased patient dependency on subsequent health care services, and decreased overall medical costs [
98].
Open abdomen
An open abdomen (OA) procedure is the best way of implementing re-laparotomies. The role of the OA in the management of severe peritonitis has been a controversial issue.
In 2007, a randomised study compared open and closed abdomens for the “on demand re-laparotomy” group in the treatment of severe peritonitis. The study was prematurely terminated following the treatment of 40 subjects due to a significantly higher mortality rate in the open abdomen group compared to the temporarily closed abdomen group (55% vs. 30%). OA procedures were performed using only non-absorbable polypropylene mesh [
99].
Although guidelines suggest not to routinely utilize the open abdomen approach for patients with severe intra-peritoneal contamination undergoing emergency laparotomy for intra-abdominal sepsis [
100], OA has now been accepted as a strategy in treating intra-abdominal sepsis [
101].
An OA approach in severe secondary peritonitis may be required for three different reasons, often used in combination: inadequate source control, severely deranged physiology (the operation is purposely abbreviated due to the severe physiological derangement and suboptimal local conditions for healing, and restoration of intestinal continuity is deferred to the second operation, i.e. the deferred anastomosis approach) [
102], and prevention of abdominal compartment syndrome [
103‐
105].
The rationale of the OA strategy in patients with severe abdominal sepsis refers to the cytokine release that is compartmentalized in the peritoneal cavity. Inability to control or interrupt the local inflammatory response is associated with higher mortality rates in these patients. The attenuation of the local inflammatory response may be best achieved with mechanical control by reducing the load of cytokines and other inflammatory substances [
106] and by preventing their production, thus removing the source itself. Sometimes more laparotomies are required to complete source control and OA allows the surgeon to perform subsequent planned laparotomies more efficiently.
An interesting non-comparative descriptive case series [
106] studied the inflammatory response in peritoneal exudate and plasma of patients undergoing planned re-laparotomy for severe secondary peritonitis. In septic patients undergoing re-laparotomy for severe peritonitis, endotoxin, tumour necrosis factor alpha, interleukin-1 and interleukin-6 levels, were higher in the peritoneal cavity then in plasma. When patients underwent re-laparotomy, the level of those cytokines was significantly decreased in survivors.
OA management has been described in patients with intra-abdominal sepsis when a single laparotomy failed to control local inflammatory response, or the risk of organ dysfunction increased after effective drainage and debridement [
107‐
109].
In the event of massive fluid resuscitation, bowel oedema and the forced closure of a non-compliant abdominal wall may cause intra-abdominal hypertension (IAH). Uncontrolled IAH exceeding 25 mm Hg may cause abdominal compartment syndrome (ACS), which is a potentially lethal complication characterized by adverse effects on pulmonary, cardiovascular, renal, splanchnic, and central nervous system physiology [
109].
The combination of IAH and the physiological effects of sepsis, result in high morbidity and mortality rates. At present there are no definite criteria to guide the surgeon in deciding whether to use the OA strategy [
110]. The OA strategy allows surgeons to extend the concept of damage control surgery to abdominal severe sepsis.
The term damage control surgery (DCS) for trauma patients was introduced in 1993. It was defined as initial control of haemorrhage and contamination, allowing for resuscitation to normal physiology in the intensive care unit and subsequent definitive re-exploration [
111,
112].
The adaptation of damage control surgery for trauma to other areas generally is useful in those patients who are at risk to develop a similar loss of physiologic reserve with intolerance to the shocked physiological state [
113]. Similarly to the trauma patient with the lethal triad of acidosis, hypothermia and coagulopathy, many patients with severe sepsis or septic shock may present in a similar fashion. For those patients, DCS can truly be life saving. Patients progressing from sepsis through severe sepsis with organ dysfunction into septic shock, can present with vasodilation, hypotension, and myocardial depression, combined with coagulopathy. These patients are profoundly haemodynamically unstable and are clearly not optimal candidates for complex operative interventions [
114].
Abdominal closure should be temporary, and the patient is rapidly taken to the ICU for physiologic optimization. This includes optimization of volume resuscitation and mechanical ventilation, correction of coagulopathy and hypothermia, and monitoring for eventual ACS developement. Over the following 24 to 48 hours, when abnormal physiology is corrected the patient can be safely taken back to the operating room for re-operation.
An additional advantage of DCS in abdominal sepsis is the possibility to delay the bowel anastomosis [
115].
The surgical strategy for the management of patients with compromised bowel in secondary peritonitis has been usually the resection of the perforated viscus followed by primary anastomosis or a diversion. In patients with severe secondary peritonitis and significant hemodynamic instability and compromised tissue perfusion, the use of primary anastomosis is limited because of the high risk of suture/anastomotic failure, leakage, and increased surgical mortality. In these patients, it is advisable to control the source of peritoneal contamination and to perform an intestinal ostomy delaying bowel anastomosis.
In a retrospective study from Colombia, 112 patients with secondary peritonitis requiring bowel resection and managed with staged laparotomy were analyzed [
116]. Deferred primary anastomosis was used in 34 patients where the bowel ends were closed at first operation and definitive anastomoses were reconstructed at the subsequent operation following physiological stabilization in the ICU and repeated peritoneal washes until the septic source was controlled. In contrast, 78 patients underwent small bowel or colonic diversion followed by similar ICU stabilization and peritoneal washes. In both groups, the abdomens were left open at the initial operation and a Velcro system or vacuum pack was used for temporary abdominal closure. The mean number of laparotomies was four in both groups. There were more patients with colon resections in the diversion group (80% vs. 47%). There was no significant difference in hospital mortality (12% for deferred anastomosis vs. 17% for diversion), frequency of anastomotic leaks or fistulas (9% vs. 5%), or ARDS (18% vs. 31%). The authors concluded that in critically ill patients with severe secondary peritonitis managed with staged laparotomies, deferred primary anastomosis can be performed safely as long as adequate control of the septic foci and restoration of deranged physiology is achieved prior to reconstruction.
In a non-randomized study of 27 consecutive patients with perforated diverticulitis (Hinchey III/IV), the patients were managed either with sigmoid resection and primary anastomosis, or limited sigmoid resection or suture, open abdomen and primary anastomosis or colostomy at second operation 24–48 hours later, or Hartmann procedure; sigmoid resection and end colostomy [
117]. All 6 patients with primary anastomosis survived without complications, but there was an obvious selection bias. Of the 6 patients undergoing Hartmann’s procedure, one died of sepsis and 5 were discharged with stoma. In the interesting group of 15 patients with deferred anastomosis or stoma and open abdomen, 9 patients had intestinal continuity restored during the second look operation with one fatal anastomotic leakage.
In a prospective study of 51 patients with perforated diverticulitis (Hinchey III/IV) were initially managed with limited resection, lavage and TAC with vacuum-assisted closure followed by second, reconstructive operation 24–48 hours later [
118]. Bowel continuity was restored in 38 patients, in 4 protected by a loop ileostomy. Five anastomotic leaks (13%) were encountered requiring loop ileostomy (2 patients) or Hartmann’s procedure (3 patients). Postoperative abscesses were seen in 4 patients, abdominal wall dehiscence in one and re-laparotomy for drain-related small bowel perforation in one. The overall mortality rate was 10% and 35/46 (76%) of the surviving patients left the hospital with reconstructed colon continuity. Fascial closure was achieved in all patients.
Following stabilization of the patient, the goal is the early and definitive closure of the abdomen, in order to reduce the complications associated with an open abdomen [
119].
A review of the literature suggests a bimodal distribution of primary closure rates, with early closure dependent on post operative intensive care management whilst delayed closure is more affected by the choice of the temporary abdominal closure technique [
120].
Primary fascial closure can be achieved in many cases within few days from the initial operation. It would not be successful if early surgical source control failed [
121,
122].
Sequential fascial closure could immediately be started once abdominal sepsis is well controlled [
123]. In these cases, surgeons should perform a progressive closure, where the abdomen is incrementally closed each time the patient undergoes a reoperation.
Within 10 to 14 days the fascia retracts laterally and becomes adherent to the overlying fat; this makes primary closure impossible. Therefore, it is important to prevent the retraction of the myo-fascial unit.
Several materials can be used to achieve temporary closure of the abdomen: gauze; mesh; impermeable self-adhesive membrane dressings, zippers and negative pressure therapy (NPT) techniques.
The ideal temporary abdominal closure method should be able to protect the abdominal contents, to prevent evisceration, to allow removal of infected or toxic fluid from the peritoneal cavity, to prevent the formation of fistulas, to avoid damage to the fascia, to preserve the abdominal wall domain, to make re-operation easy, safe and facilitate definitive closure [
110].
The surgical options for management of the OA are now more diverse and sophisticated, but there is a lack of prospective randomized controlled trials demonstrating the superiority of any particular method.
At present, negative pressure therapy (NPT) techniques have become the most extensively used methods for temporary abdominal wall closure. NPT actively drains toxin or bacteria-rich intra peritoneal fluid and has resulted in a high rate of fascial and abdominal wall closure [
110].
A systematic review conducted in 2012 [
124] found only 11 comparative studies, including 2 randomized controlled trials (RCTs) and 9 cohort studies, examining the efficacy and safety of negative pressure peritoneal therapy
versus alternate temporal abdominal closure methods among critically ill or injured adults.
However, all studies were associated with at least a moderate risk of bias and significant clinical heterogeneity, the authors concluded that there was insufficient evidence to support the preferential use of negative pressure peritoneal therapy after damage control laparotomy.
Animal data suggest that OA techniques employing constant negative pressure to the peritoneal cavity may remove inflammatory ascites, reduce the systemic inflammatory response, and improve organ injury and potentially outcomes [
125].
This method is still associated with high morbidity and high incidence of ventral hernia formation in surviving patients caused by difficulties in definitive closure of the abdominal wall after prolonged application of NPT but it could be a highly promising method in the management of patients with increased IAP and severe sepsis due to severe peritonitis [
126].
A systematic review published in 2009 [
127] investigated which temporary abdominal closure technique is associated with the highest delayed primary fascial closure (FC) rate.
No comparative studies were identified. 51 articles were included. The techniques described were vacuum-assisted closure (VAC; 8 series), vacuum pack (15 series), artificial burr (4 series), Mesh/sheet (16 series), zipper (7 series), silo (3 series), skin closure (2 series), dynamic retention sutures (DRS), and loose packing (1 series each).
These results suggested that the artificial burr and the VAC were associated with the highest FC rates and the lowest mortality rates.
Other techniques used for progressive FC include a combination of NPT with a temporary mesh sutured to the fascial edges. The mesh is tightened every few days, until the fascial defect is small enough so the mesh can be removed and the fascia closed primarily.
In 2012, a retrospective analysis evaluating the use of vacuum-assisted closure and mesh-mediated fascial traction (VACM) as temporary abdominal closure was published [
128]. The study compared 50 patients treated with (VACM) and 54 using non-traction techniques (control group). VACM resulted in a higher fascial closure rate and lower planned hernia rate than methods that did not provide fascial traction.
Occasionally, abdominal closure is only partially achieved, resulting in late development of large, debilitating hernias of the abdominal wall which will eventually require complex surgical repair. In these cases, delayed repair or use of biological meshes has been proposed [
129].
Another option, if definitive fascial closure is not possible, is closure of the skin only and subsequent management of the eventration by a deferred abdominal closure with synthetic meshes after hospital discharge [
127].