Peritonitis

Peritonitis is a clinical syndrome characterized by an inflammatory response to irritation of parietal serous membranes that line the abdominal cavity and the visceral membranes that cover the abdominal viscera, and it is a common initiator of the systemic inflammatory response syndrome [1]. It is most commonly secondary to diseases or trauma that result in bacterial contamination of the abdominal cavity in both dogs and cats. Introduction of microbes by any means can induce septic peritonitis, whereas non septic inflammation, or aseptic peritonitis, may be induced by neoplastic invasion, blood, urine extravasation, bile leakage, and irritating acid by-products of pancreatitis [2-4]. The latter nonseptic conditions may become septic (i.e., bowel obstruction) if damage to bowel allows transmural movement of bacteria.3 Primary peritonitis is not common in small animals and is usually attributed to corona viral infection causing feline infectious peritonitis in cats, or to hematogenous infection if the source of peritonitis is not known in either dogs or cats [5]. The surface area of these inflamed membranes can be as much as 150% percent of the surface area of total body surface area [6], so disorders that affect the abdominal cavity in either a generalized or local region can have a profound affect on the clinical health of the dog or cat.

Anatomy

The peritoneum is a serous membrane made up of mesenchymal cells supported by a stroma of collagenous and elastic fibers, called the transverse fascia [6]. Other cells and substances populate this stroma and include macrophages, lymphocytes, mast cells, adipose cells, and glycosaminoglycans. Special lymphatic collecting vessels and lacunae, especially common on the visceral surface of the diaphragm, are responsible for clearing fluid and particles from the peritoneal cavity [6,7], but are also capable of increasing in diameter during peritonitis [8]. The peritoneal cavity normally has a small amount of serous fluid that serves a lubricating purpose. This fluid is produced by the mesenchymal cells [6]. The peritoneal membrane has been characterized as a semipermeable membrane that is capable of absorptive and exudative functions [2]. Circulation of normal peritoneal fluid occurs caudal to cranial (toward the diaphragm). Investigative dye studies show clearance times varying from 19 to 72 minutes, depending on a cranial or caudal location of the dye injection, respectively [6,9]. After absorption by diaphragmatic lymphatics, lymph is carried to mediastinal lymph nodes, to the thoracic duct, and finally to the systemic circulation. Omental lymphatics provide a route for drainage and aggregations of cells ("milky spots"), which include neutrophils, macrophages, and lymphocytes that are important to defense mechanisms. Peritoneal-associated lymphoid tissues are also capable of immunoglobulin production [6].

The peritoneal cavity has a normal positive pressure, which varies from 1.5 to 5.5 mm Hg, which can be measured indirectly using a transurethral urinary bladder catheter. Increases in peritoneal pressure can result in decreased abdominal compartment compliance, which can cause cardiovascular, respiratory, and abdominal organ dysfunction. Peritoneal pressures greater than 15 mm Hg can be associated with tachycardia, hypertension, increased systemic vascular resistance, decreased cardiac output, decreased mesenteric blood flow, decreased intestinal mucosal blood flow, and increased bacterial translocation [6]. Intra-abdominal pressures are increased after ovariohysterectomy and other causes of abdominal distention. Abdominal pressures greater than 22 mm Hg may require surgical decompression, especially in oliguric animals [10].

Cytologic characteristics of normal peritoneal fluid include mostly macrophages, mesothelial cells and lymphocytes, and a protein content of less than 3 g/dl. Peritoneal fluid lacks fibrinogen, does not clot, and has antibacterial activity as a result of its complement and opsonizing fibronectin components [6].

Etiology of Secondary Peritonitis

Gastrointestinal microorganisms gain access to the peritoneal cavity by leakage from perforation or other loss of bowel integrity and are the most common cause of secondary peritonitis [6,11]. Peritonitis may follow abdominal surgery and, when postoperative peritonitis occurs, the mortality rate is high [11]. Recently, preoperative peritonitis has been shown to be a risk factor for developing anastomotic leakage [12]. Bacteria and inflammatory cells produce collagenases, which decrease the collagen content of the intestinal wall and impair the strength of the anastomosis [12]. Gram-negative aerobes and anaerobes with an admixture of fluid and ingesta from the bowel result in initiation of inflammation from bacterial, chemical, and foreign materials. Leakage from bowel may occur from mechanical perforation, neoplastic invasion, foreign materials, sharp and blunt trauma, vascular disruption causing ischemia and necrosis, dehiscence of surgical incisions, and drug-induced lesions such as corticosteroid-induced ulceration or colonic perforation. Gastrointestinal perforation typically results in a plural population of bacteria within peritoneal fluids. A factor that determines the number and type of bacteria that escape into the peritoneal cavity is which region of gastrointestinal tract perforates. The more distally the site of perforation is located, the higher are the total bacteria count and the percentage of anaerobic microorganisms, as well as the incidence of mortality [6]. Two bacteria predominate with intestinal perforation: E. coli and Bacteroides fragilis. The aerobic and anaerobic characteristics of these organisms are thought to be synergistic. Endotoxin produced by E. coli is commonly integral in early mortality, and concomitant infection with Bacteroides fragilis is thought to enhance the lethal potential of E. coli [13,14]. Alpha-hemolysin, an exotoxin produced by E. coli, is thought to promote infection within the peritoneal cavity, as it is toxic to mammalian cells, and it alters the character of intraperitoneal fluid by lysing red blood cells and having detrimental effects on peritoneal leukocytes [14,15]. This exotoxin enhances the ability to recover E. coli and Bacteroides fragilis from peritoneal fluid, and it increases mortality [14].

Rupture of a septic organ such as the pancreas, prostate, uterus, gallbladder, or urinary bladder, and rupture of intra-abdominal abscesses (i.e., liver) usually result in single population of bacteria. The fluid content varies depending on the septic organ that ruptures. Peritoneal dialysis is commonly complicated by septic peritonitis, usually a result of tubing contamination during fluid exchanges or gastrointestinal perforation [16]. Peritonitis is rarely secondary to parasitic infections. Toxoplasma gondii infection in cats, Porocephalus crotali, and mesocetoides species in dogs have been identified [17-19].

Secondary peritonitis that occurs in the absence of an infectious pathogen is termed aseptic peritonitis. Chemical agents (usually endogenous such as urine or bile), foreign materials such as starch or surgical sponges, mechanical irritation, and neoplasia may result in aseptic peritonitis. Surgical exposure and manipulation, including exposure to air during surgery or laparoscopic inflammation will cause peritoneal inflammation. Granulomatous peritonitis may result from glove powders, including starch, from talcum powder, and from silicone-based powders. Hypersensitivity to starch or corn is the proposed mechanism [6]. Sterile urine and bile within the peritoneal cavity does not result in significant inflammation of the peritoneum unless bacteria gain access to the fluid [17,20,21]. Sterile bile effusions have been reported for as long as 30 days [21]. Uroperitoneum usually has a more significant metabolic impact on the health of the animal; bacterial contamination from urine is not common. Sterile urine may remain in the peritoneal cavity for long periods without peritonitis if it is sterile [22]. If uroperitoneum is contaminated by bacteria, the need to correct the problem surgically is more urgent, whereas if the urine is sterile, urine diversion via peritoneal drainage can be performed until the patient is stable. Other chemicals that are iatrogenically placed into the abdominal cavity, such as antibiotics, barium, povidone-iodine solutions, and iodinated contrast agents cause varying degrees of peritoneal inflammation [6].

Sclerosing encapsulating peritonitis, a chronic condition that results in the abdominal organs being encased in "cocoon-like" layers of collagenous connective tissue, has been reported in small animals [6]. This condition is nonseptic, with peritoneal fluid containing red blood cells, macrophages with phagocytized red cells, and fibroblasts. The etiology of this condition is uncertain [23,24].

Inflammation of Peritonitis

The peritoneum utilizes immunologic mechanisms, absorptive function, and the ability to localize infectious processes to defend its surfaces. When peritoneal contamination or injury occurs, an immediate inflammatory response occurs beginning with activation of complement (C3a and C5a), and the influx of neutrophils. Mast cells and basophils degranulate, promoting vascular permeability and the elaboration of opsonins and complement. Opsonization of organisms, cell lysis and clearance of immune complexes is potentiated by complement. Phagocytosis is enhanced, and immunoglobulins are produced by peritoneum-associated lymphoid tissues [25]. Mast cells, neutrophils, macrophages, lymphocytes, and mesenchymal cells participate in release of cytokines, which results in further cell recruitments. Prostaglandin synthesis results from arachidonic acid metabolism stimulated by interleukin 1β and tumor necrosis factor-α. Such cytokines, with interleukin 8, enhance neutrophil emigration. The elevation of proinflammatory cytokines, such as tumor necrosis factor-α, and interleukins 8 and 6 are proportional to the severity of the clinical response [6]. Exudation of peritoneal fluid results from increased permeability of peritoneal capillary networks caused by mast cell histamines and prostaglandins. This peritoneal fluid exudation provides a large volume of fluid containing complement, immunoglobulins, clotting factors, and fibrin. Fibrin clearance is decreased because the fibrinolytic system is inactivated by inflammation, and clumps of fibrin can occlude peritoneal lacunae. Bacteria may be protected from the inflammatory response in regions of fibrin deposit; however, fibrin deposits are prerequisite to fibrous adhesion formation, which help localize regions of infection [6,26].

The consequence of exudation of fluid from the vascular space into the peritoneal cavity is hypovolemia and hypoproteinemia. Hypovolemia and hypoproteinemia are exacerbated when ileus occurs secondary to sympathoadrenergic reflex inhibition [27] and fluid becomes sequestered in the bowel lumen. Translocation of bacteria is enhanced when intestinal motility is altered; therefore, in nonseptic types of peritonitis the associated ileus may allow transmural bacterial contamination of the peritoneal cavity [6]. Ileus may also occur from conditions such as ischemia or chronic distention from obstruction. As the volume of fluid in the peritoneal cavity continues to increase, loss of diaphragmatic compliance occurs and ventilation is compromised. When severe, hypoxemia and respiratory acidosis can result. Increasing volumes of peritoneal fluid eventually increase intraabdominal pressure, which may reduce venous return from abdominal capacitance vessels and negatively impact cardiac output. Acute renal failure will eventually result from decreased renal perfusion. Septic peritonitis has been associated with septic hepatopathy; intrahepatic cholestasis, icterus, and elevation of serum bile acids and liver enzymes [28]. Disseminated intravascular coagulation resulting in microembolization of parenchymal organ blood supply exacerbates insults to major organs from hypovolemia and hypoxia. Multiorgan dysfunction syndrome (MODS) may result from cell dysfunction and is mediated by cytokines.

Sepsis commonly occurs when peritonitis is secondary to bacterial contamination. Secondary effects of the bacterial pathogens and their by-products complicate the metabolic alterations already described. Experimentation has shown sepsis to be associated with a marked increase in peripheral oxygen demand, elevated levels of plasma insulin, glucagon, and catecholamines (the hyperdynamic state) [29,30]. Endotoxin potentiates levels of proinflammatory cytokines, complement, and products of arachidonic acid metabolism, by stimulation of the innate immune response [6]. Anaerobic organisms produce exoenzymes that make them particularly invasive to tissue, causing necrosis, suppuration, adhesion, and abscess formation [6].

Substances that Potentiate Peritonitis

Substances that augment or potentiate local or systemic inflammatory responses in peritonitis or that worsen the prognosis for recovery in bacterial peritonitis are termed adjuvants of peritonitis [29]. Bile salts, gastric mucin, hemoglobin, and barium are recognized as adjuvants that enhance the virulence of contaminating bacteria [6,31]. Phagocytosis is inhibited by gastric mucin because of a heparin-like anticomplement effect and by hemoglobin, which disrupts phagocytic cell chemotaxis. Hemoglobin also interferes with phagocytosis, intracellular killing [29], and lymphatic clearing mechanisms [26,32]. Bile salts lyse red blood cells and release hemoglobin and alter cellular adhesion mechanisms by lowering surface tension [33]. Another adjuvant effect is related to peritoneal fluid volume. Experimentally, incremental increases in volumes of fluid injected intraperitoneally while keeping the bacterial inoculums unchanged resulted in slowed bacterial clearance, increased bacterial proliferation, and increased mortality [34]. Barium and intestinal-content contamination resulted in a higher mortality rate than intestinal content contamination alone in an experimental study [35]. Despite this finding, barium is still used in upper gastrointestinal studies because it provides better radiographic detail. If leakage occurs, immediate exploratory surgery is performed to resolve the gastrointestinal perforation and to remove extravasated barium and foreign material by thorough lavage.

Clinical Signs of Peritonitis

It is common to think of clinical signs referable to peritonitis as acutely painful with vomiting, fever, and shock, but peritonitis in dogs and cats can present with a wide variation in clinical signs [6]. Anorexia, vomiting, malaise, depression, fever, weakness, and an abdominal pain response are common. Fever is not always present and rectal temperature can vary from hypothermic to hyperthermic. Fever may occur in either septic or aseptic peritonitis [36]. Dogs generally exhibit an abdominal pain response, however one third of cats with peritonitis do not exhibit an abdominal pain response [37]. Abdominal pain may be exhibited by a "praying position", which may provide pain relief in some dogs (also not a consistent observation). Varying degrees of abdominal fluid accumulation may or may not be detectable during physical examination. Large volumes of effusion may cause respiratory compromise. Sequestration of fluid in the peritoneal cavity and the intestinal lumen from ileus can quickly result in marked dehydration when accompanied by vomiting. In cats and dogs, peritoneal effusions have been associated with hyponatremia and hyperkalemia in the presence of normal adrenal gland function [38]. Auscultation of the abdomen may reveal the absence of borborygmi, consistent with ileus. Vital signs that may be suggestive of a hyperdynamic systemic inflammatory response syndrome include injected (brick red) mucous membranes, rapid capillary refill time, tachycardia, strong quick pulses, and pyrexia. As peritonitis progresses and hypovolemia worsens, tachycardia becomes severe with weak pulses, prolonged capillary refill time, pale membranes, and hypothermia (hypovolemic shock). In cats, bradycardia and hypothermia, pale mucous membranes, signs of diffuse abdominal pain, weak pulses, anemia, hypoalbuminemia, and icterus are indicative of severe sepsis [39]. Clinical signs may also reflect the organ system involved: icterus with septic cholecystitis and perforation; a vaginal discharge with ruptured pyometra; dysuria or pyuria with prostatic abscessation or septic prostatitis. Purulent drainage associated with abdominal pain after gastrointestinal surgery can be associated with dehiscence and leakage of bowel content.

Diagnosis

Many diagnostic tests contribute to and support the diagnosis of peritonitis, but cytologic examination of peritoneal fluid is the most important examination. The finding of degenerative neutrophils with intracellular bacteria is diagnostic for septic peritonitis. Culture (and susceptibility test) confirms bacterial infection [40]. A recent study recommended culturing nonseptic fluid, as lack of cytologic evidence for sepsis did not mean bacteria would not be cultured [41]. Often the volume of peritoneal fluid makes obtaining peritoneal fluid by abdominocentesis relatively simple, but ultrasound can guide aspiration when the fluid volume is low. In the absence of ultrasound guidance, a "four-quadrant" tap may be done. Diagnostic peritoneal lavage may also be used to "wash" peritoneal surfaces and collect fluid for examination; it is indicated when septic peritonitis is suspected despite low volumes or absence of abdominal fluid [6]. Exploration of the abdominal cavity is indicated if cytologic examination reveals degenerative or toxic neutrophils with phagocytized bacteria, free bacteria (be careful of stain contaminants), or plant material. Abdominal fluid obtained after uncomplicated gastrointestinal anastomosis surgery typically had non degenerate neutrophils [42]. Neutrophil counts vary widely; cytologic characteristics are considered the most important criteria and should be used in conjunction with peripheral white blood cell counts. Peripheral blood WBC can vary from neutrophilia to neutropenia when neutrophilic consumption exceeds bone marrow production [36]. Culture of aerobic and anaerobic organisms and susceptibility testing is done.

Peritoneal fluid and serum in septic peritonitis have been compared, as bacterial isolation is not temporally realistic versus cytologic diagnosis. In this study, a presurgical white blood cell count of more than 2000 cells/dl and postsurgical white blood cell count of more than 9000 cells/dl was indicative of peritonitis in general. When blood and peritoneal fluid glucose were compared, a concentration difference of more than 20 mg/dl differentiated septic peritoneal effusions from nonseptic effusions in dogs and cats [40]. Lactate production occurs as a result of neutrophilic glycolysis and bacterial metabolites from an anaerobic microenvironment in peritoneal fluid with a resultant decrease in the pH of peritoneal fluid [43]. Septic effusions in cats were found to have a lower pH, but dogs did not show this trend [40]. Differences in blood to fluid pH were found to be insignificant in dogs and cats, blood to fluid glucose significant in dogs and cats, and blood to fluid lactate insignificant because of low numbers. Although not statistically significant, a blood to fluid lactate difference of less than -2.0 mmol/L was 100% sensitive and specific for septic peritoneal effusion in 7 dogs [40]. Blood to fluid glucose difference was more sensitive than glucose alone [40]. Another investigation in 19 dogs and 18 cats, showed dogs with septic peritoneal effusions to have peritoneal fluid lactate concentrations of more than 2.5 mmol/L and to have peritoneal fluid lactate concentrations that were higher than blood lactate concentrations (a negative blood to peritoneal fluid lactate difference); however, similar tests were not found to be accurate in cats [43]. It has been hypothesized that, because cats are deficient in glucokinase, they may have a tendency toward anaerobic metabolism with higher concentrations of lactate in peripheral blood [43].

Suspicion of peritoneal fluid may occur during physical examination if a fluid wave is palpable and survey radiographs of the abdomen demonstrate loss of serosal detail (a ground glass appearance). In the absence of recent abdominal surgery, the appearance of free air in the abdominal cavity is indicative of gastrointestinal rupture [41]. Studies have shown gastrointestinal perforation to be the cause in 77% of 34 animals with pneumoperitoneum [41], and in 74% of 54 animals with pneumoperitoneum but without history of penetrating trauma [44]. Residual free air can occur after abdominal surgery for as long as 30 days [41], or from diagnostic procedures that utilize gas or air. Emphasematous change within abdominal organs may show air density as well, and be associated with peritonitis [6]. Ultrasonographic examination can reveal fluid and may be able to localize the potential source of leakage if a parenchymal organ is the source of septic purulent fluid. Corrugated small intestine was noted in 4 of 24 dogs with peritonitis, but was more commonly associated with pancreatitis (12 of 24 dogs) [45]. Pleural effusion may occur concomitant with peritonitis and is considered a poor prognostic sign [46]. Imaging techniques such as CT and MRI may provide additional information prior to exploratory surgery, but ultrasonography provides a simple, practical, and widely available tool.

Complete blood count, serum chemistry evaluation, and coagulation profiles are preferentially done to assess the severity of disease. Biochemical analysis of peritoneal fluid can be helpful if uroperitoneum or bile leakage is suspected. Creatinine levels in peritoneal fluid greater than serum levels support a diagnosis of uroperitoneum; however, in icteric animals, reagent-based tests for bilirubin (i.e., urine reagent strips) are not accurate [47]. Determination of bilirubin concentration has been shown to be 100% effective in diagnosing bile leakage prior to exploratory surgery. It was consistently two and one-half times the serum concentration of bilirubin when bile leakage was occurring [21]. Hypoglycemia is often present during sepsis [36]. A glucose concentration in peritoneal fluid of less than 50 mg/dl was 100% specific for bacterial peritonitis when 55 cases of nonseptic abdominal effusion were compared with 16 cases of septic abdominal effusion [48].

Management of Septic Peritonitis

When making a diagnosis of septic peritonitis supportive treatment is initiated typically with aggressive intravenous fluid resuscitation to restore hydration and to improve perfusion. Intravenous crystalloids are administered initially with the goal of achieving a urine output of 1 to 2 ml/kg/hour with a central venous pressure between 0 and 5 cm H2O. Monitoring of central venous pressure can help tailor fluid therapy to achieve volume expansion and avoid fluid overload. Synthetic colloid administration or blood products may be appropriate, depending on the results of blood and serum evaluations. As fluid resuscitation progresses, serial testing of CVP, serum albumin, colloid osmotic pressure, acid base status, electrolytes, coagulation parameters, and blood pressure aid therapeutic decisions. Preoperative analysis can help to relate cardiovascular parameters to the effects of fluid administration. Fluid rates of 10 to 12 ml/kg/hour may be required to assure maintenance of blood pressure after surgery [6]. Antimicrobial therapy is initiated as soon as peritoneal fluid specimens are cultured and generally involves a combination of an aminoglycoside and a parenteral drug that is effective against anaerobes; however, this greatly depends on individual clinician's preference [6]. Administration of antimicrobials, such as cefoxitin, that have broad spectrum and good activity against anaerobes, simplifies initial antimicrobial therapy. Antimicrobial therapy may be altered as soon as susceptibility tests are returned. Exploration and resolution of the source of bacterial contamination are required for successful treatment of septic peritonitis. Thorough lavage of the peritoneal cavity to dilute contaminants is believed to be important by most surgeons, but a recent investigation questions if evidence-based support exists for lavage [49]. Antimicrobials are not used in the lavage fluid during surgery as parenteral antimicrobials reach sufficient (therapeutic) levels in peritoneal fluid during peritonitis [50], and adverse affects such as chemical irritation, adhesions, and delayed anastomotic healing are avoided [51]. In addition, use of antimicrobials in lavage fluids has not been shown to provide significant benefit over lavage alone. After correcting the source of bacterial contamination and thorough lavage, consideration is given to providing a gastric or enteral route of alimentation, so nutritional support can be provided in the early postoperative period. Corticosteroid and nonsteroidal use in septic peritonitis is controversial and is not routinely done, as there is no proven benefit [36].

Drainage of the inflamed peritoneal cavity is the decision that seems most controversial. Closure of the abdominal cavity without drainage, sump drains for continued drainage (with or without intermittent lavage), open peritoneal drainage, and vacuum-assisted peritoneal drainage are used currently, although the latter technique is in its infancy in both human and veterinary medicine [52-55]. Open drainage allows the fluid to be removed from the abdomen in as little as 6 hours, whereas sump drainage requires 24 to 48 hours [6]. A retrospective investigation of 36 dogs and 6 cats, comparing open peritoneal drainage and primary closure techniques, found an overall survival rate of 71% with no significant difference in survival between groups. In the open drainage group, however, patients received more plasma and blood, more animals had jejunostomy tubes and longer intensive care unit hospitalization (a mean of 6 days versus 3.5 days for the primary closure group) [55]. Prospective clinical trials comparing closed and open drainage techniques are not available in the veterinary literature.

Primary closure of the abdominal incision after exploration for septic peritonitis has been reported to have a mortality rate of 46% [53]. Although prospective comparison of open and closed methods of treating septic peritonitis has not been done, retrospective comparison of mortality rates in individual studies shows mortality rates to be similar, with gastrointestinal leakage being the most common cause and having the highest mortality [52]. A prospective randomized study that categorized septic peritonitis as gastrointestinal, biliary, and nonbiliary nongastrointestinal (i.e., uterus, prostate, renal) such that large numbers of animals managed by open and closed methods in each category would contribute to determination of when open peritoneal drainage is warranted and when primary closure is sufficient. In the author's opinion, septic peritonitis that originates from parenchymal organs such as the uterus, prostate, and kidney tends to have peritoneal contamination that is easily removed and diluted by large volume lavage. In situations of gastrointestinal content leakage, the decision for open peritoneal drainage is made based on judging the adequacy of lavage to dilute and remove foreign material and exudates and the extent of surface involvement (local or diffuse).

Septic peritonitis treated by primary closure and abdominal drains has not achieved strong support because drainage has been ineffective and because it benefits only a local region of the peritoneal cavity, undergoes premature partial occlusion by omentum, and can be complicated by ascending nosocomial infection. Bacterial contamination via an abdominal drain has been shown to occur in as little as 24 hours [56]. Sump drains allow better drainage than non-sump drains but also allow potential contamination because air is pulled into the peritoneal cavity if not filtered. Sump-Penrose drains were completely encased by omentum and omental adhesions in 96 hours in normal dogs [54]. Both the sump-Penrose drains and open peritoneal drainage in normal dogs caused a local inflammatory response [54]. Intermittent peritoneal lavage with a Parker peritoneal dialysis cannula has also been used, and an average of 91.4% of the infused lavage volume was recovered. Closed suction drains have also been used successfully to treat generalized peritonitis in dogs and cats without clinically important complications [57,58].

Open peritoneal drainage has resulted in mortality rates varying from 22% to 48% [52,54,55]. Open peritoneal drainage provides the best and most complete drainage of the peritoneal cavity, essentially leaving it as an open wound or abscess would be treated and maintaining a microenvironment within the cavity less favorable to anaerobic bacteria. The efficacy of open peritoneal drainage is reported to be due to enhanced removal of bacteria, foreign material, and exudates (including inflammatory mediators) [59]. Gross appearance of the wound, cytologic examination of the fluid, and the condition of the patient are factors that contribute to the optimum timing of abdominal closure. Reexploration of the abdomen may be performed if degenerative neutrophilic inflammation or bacterial contamination persists or returns. Bacterial culture is done prior to closing the abdomen; in previous studies, as high as 40% of cases had different bacteria isolated at closure than at initial exploration [6]. The most common complications of open peritoneal drainage include hypoproteinemia, hypoalbuminemia, anemia, and nosocomial infection [6]. Open peritoneal drainage in humans has been shown to have significantly more complications and no advantage over a primary closure technique when compared in a prospective clinical trial [60]. A study of 239 patients showed 31% mortality with closed technique and 44% mortality with open techniques [61]. The suggestion has been made that open peritoneal drainage is possibly enhanced by the animal's posture [6].

Vacuum-assisted closure (VAC) is a new technique in both human and veterinary medicine and has application in acute and chronic wounds [62,63]. Vacuum-assisted peritoneal drainage is in its infancy, but is currently being evaluated in human patients and in a few veterinary centers. The technique has been shown to accelerate wound healing by increasing local blood flow, by reducing bacterial load, and by stimulating growth of granulation tissue [64]. In humans, use of temporary vacuum-assisted closure is being used for local and for generalized peritonitis. Techniques have been developed for using VAC to treat anastomotic leakage after rectal resection [64]. Vacuum-assisted wound closure has been reported to augment open abdominal wall repairs where human acellular dermal matrix is used to close open abdomens that cannot be closed by local tissue mobilization [65]. Use of VAC has also been successful in managing post-laparotomy wound dehiscence where compromised wound healing was thought to play a complicating role [66]. Technique modifications are being developed to allow treatment of local intra-abdominal infections without dissemination of infection to other areas of the abdominal cavity [67].

Supportive Care

Septic peritonitis results in massive protein and electrolyte losses in an animal that is unlikely to eat soon after surgery. Failure to provide nutritional support results in protein-energy malnutrition, which depletes energy stores, delays wound healing, impairs immunocompetence, and can result in weakness and eventually organ failure [68]. Early enteral nutrition is beneficial to enterocytes and also has been shown to decrease bacterial mural translocation, preserve or increase gastrointestinal blood flow, prevent ulceration, increase IgA concentration and stimulate other immune system defenses, and enhance wound repair [68]; therefore, a judgment as to what technique to use is important during surgery. Jejunostomy tubes allow direct infusion (as continuous rate infusion) of special diets into the small intestine; however, the complication rate with this technique is high, varying from 17.5% to 42% [6]. Esophagostomy tube placement and gastrostomy tube placement are alternatives to jejunostomy. Other noninvasive tube placement techniques include nasoesophageal and nasoenteric feeding tubes. Parenteral nutrition may be done in lieu of enteral feeding techniques and is capable of maintaining serum protein concentrations. Transfusion therapy is important to overall management as indicated by results of serial examination of hematocrit, albumin, and coagulation parameters, and management of pain with analgesics before and after surgery.

Prognosis

Survival rates for generalized peritonitis have varied from 52% to 79%. As has been noted [6] more recent survival rates are improved and are likely related to improved diagnosis and preoperative and postoperative management. Septic bile peritonitis has been shown to be particularly lethal; only 27% of animals survived in one retrospective investigation [21]. In the same investigation, all six animals with sterile bile leakage survived [21].