Septic peritonitis is a relatively common reason for admission to our hospital and represents an important cause of morbidity and mortality. These cases can be challenging to stabilize before and during surgery and are often hospitalized for several days following definitive surgery.
Peritonitis can be classified as primary or secondary, acute or chronic, septic or aseptic, and diffuse or localized. The majority of surgical cases presenting to our hospital are secondary, acute, septic and diffuse. Primary peritonitis is the result of hematogenous spread of microorganisms, the most common example being feline infectious peritonitis caused by a coronavirus infection. Infectious secondary peritonitis can be caused by gastrointestinal or urogenital leakage (perforation, rupture, surgical dehiscence), hepatobiliary leakage (rupture, abscess, torsion etc.), intraabdominal abscess (pancreas, spleen etc.), surgical contamination or penetrating wounds. Non-infectious secondary peritonitis is commonly seen with neoplasia, pancreatitis, and sterile bile peritonitis or sterile uroabdomen.
Clinical signs associated with peritonitis (Table 1), sterile or septic, vary widely between patients and none are pathognomonic. Diagnosis can be suspected based simply on history and physical exam findings. Hematology, serum chemistry and abdominal imaging are the foundation of any investigation for acute abdomen. Plain radiographs may be challenging to interpret especially if the patient had undergone recent surgery or abdominocentesis. Some peritoneal fluid and some air is expected after any abdominal procedure, so the mere presence of fluid and air on radiographs or an abdominal ultrasound of a post-operative patient is not necessarily indicative of septic peritonitis. If the patient has NOT had recent abdominal surgery, free air in the abdomen is strongly supportive of gastrointestinal leakage. Free air can be seen easily by an experienced ultrasonographer and can also be visualized on plain radiographs. The classic view to identify free abdominal air is a horizontal beam view with the patient in lateral recumbency, but pneumoperitoneum can be seen on standard VD and lateral views as well. Contrast study radiographs, e.g. barium study, are generally contraindicated if septic abdomen is suspected. Barium leakage into the abdomen is particularly difficult to remove from the abdomen even with copious lavage and increases the mortality rate over frank gastrointestinal leakage. If a contrast study is necessary, iodinated water-soluble contrast agent should be used.
|Anorexia and/or depression|
|Vomiting and/or diarrhea|
|Abdominal pain and distension (+/- fluid wave)|
|Tachypnea and respiratory distress|
|Serosanguinous and/or purulent drainage from previous surgical site|
|Lack of borborygmi|
|Progressive signs of shock (tachycardia, poor pulse quality, prolonged CRT, pale mm, hypothermia)|
Chronic peritonitis patients usually have copious abdominal effusion making collection relatively straightforward. However, in early peritonitis or more localized disease abdominocentesis via a four quadrant approach or diagnostic peritoneal lavage may be necessary to collect a representative sample. Once fluid is collected the sample should be divided immediately into culture and analysis (Table 2). It is critical that samples are analyzed as quickly as possible after collection. In human cases of suspected septic peritonitis, samples are immediately transferred to blood culture tubes at the time of collection which increases the return of a positive culture from 42% (when transfer of the sample is delayed for even a short period of time) to 91% and is considered standard of care for sample handling. If the sample cannot be divided, priority should be given to cytology, however in companion animals cytology is 87% or less accurate in generating a diagnosis of septic peritonitis. The presence of intracellular (+/- extracellular) bacteria and degenerative neutrophils indicates septic peritonitis and warrants immediate surgical exploration. If intra- and extracellular bacteria are seen on cytology, a Gram stain can be done to help guide antibacterial treatment until culture results are returned. Glucose concentration can also be rapidly evaluated on abdominal fluid (baring gross hemorrhage in the fluid). Multiple studies have shown that patients with septic peritonitis consistently had significantly lower glucose concentration in abdominal fluid compared to peripheral blood glucose. In dogs, blood-to-fluid glucose difference greater than 20mg/dL was 100% sensitive and 100% specific for a diagnosis of septic peritonitis. In cats this magnitude blood-to-fluid glucose difference was 86% sensitive and 100% specific. Blood-to-fluid lactate difference was also investigated but found not to be as useful in cats or dogs in diagnosing septic peritonitis. Peripheral blood lactate concentration has been useful in generating a prognosis for dogs with septic peritonitis. Dogs with plasma lactate concentration >2.5 mmol/L on admission was associated with mortality as was the inability to normalize plasma lactate concentration within 6 hours of admission.
|Fluid analysis (TP, cell count/μL, specific gravity)|
|Glucose mg/dL (and blood-to-fluid difference)|
Antimicrobial treatment should be initiated immediately upon diagnosis of septic peritonitis. Ideally, antibiotics are given intravenously after collection of abdominal fluid for culture and sensitivity. If collection of fluid and transfer to culture media must be delayed, antibiotic coverage should be started anyway to help prevent further bacterial translocation and hematogenous spread thus reducing later complications. Bactericidal drugs effective against both gram-positive and gram-negative aerobes and anaerobes are recommended (Table 3). Intraperitoneal administration of antibiotics or antiseptics is not recommended and is associated with further morbidity and mortality.
|Ampicillin + gentamicin + metronidazole|
|Ampicillin + enrofloxacin + metronidazole|
|Ampicillin + amikacin|
|Cefazolin + amikacin/gentamicin|
|Extended spectrum cephalosporin (cefoxitin or cefotaxime or ceftazidime) (single agent)|
|*Aminoglycocide use only in normovolemic patients without renal compromise|
Patients should be stabilized prior to surgery with appropriate fluid and electrolyte replenishment, pain medications, heat support and oxygen supplementation. If possible, patients should be normothermic and normotensive prior to anesthetic induction but this may not be possible in severely ill patients. Rapid surgical intervention may be the best route to making meaningful progress in stabilizing some cases.
Surgical goals include correction of the source of contamination or infection, reduction of bacterial load, removal of foreign material and inflammatory products (cells and mediators) and prevention of persistent or recurrent abdominal infection. A thorough exploration of the abdominal cavity through a generous xyphoid to pubis incision with close inspection of any previous surgical repairs is mandatory. Immediate suctioning of septic fluid during exploration is optimal. Once the source of contamination is located and isolated, an initial lavage and suction of the abdomen is recommended to slow further absorption of bacteria and inflammatory products. More than half of reported cases in the veterinary literature (dogs and cats) are pets that had abdominal surgery within 14 days of presenting for septic abdomen, however it is unknown what percentage of those cases had a septic abdomen at the time of the first surgery. It is extremely difficult to discriminate between a septic abdomen that is secondary to an intestinal repair (e.g. enterotomy or anastomosis) that leaked versus an intestinal repair that leaked secondary to a septic abdomen. Healing of tissues in the face of septic peritonitis is impaired due to several mechanisms which should influence the surgeon’s choice of repair and materials. Peritonitis results in enhanced proteolytic activity that degrades collagen and extracellular matrix. This process puts enterotomy, cystotomy and intestinal anastomotic closures at high risk for failure. Suture selection should be monofilament for all applications in a septic peritonitis patient. Suture material should be absorbable but with a profile that will maintain holding strength in an inflammatory environment. For instance, a fast-absorbing monofilament sutures such as Monocryl or Caprosyn, which are excellent in tissues where healing will be uncomplicated and inflammation is minimal, would lose strength much faster than PGA (polyglycolic acid) or PDS (polydioxanone) in a septic environment, potentially prior to tissue healing and with catastrophic results. Suture material selection applies not only to visceral closure but also to the linea, subcutaneous and skin closures. Braided suture should be avoided as the interstices between suture filaments are a perfect environment for bacterial colonization resulting in persistent infection. Stainless steel suture can be used for septic abdomen cases for closure of the linea, however it is still at risk for biofilm and recurrent infection.
Reinforcing enteric closures via omental wrapping or serosal patching is recommended for septic peritonitis patients. Both procedures are designed to bring additional blood supply to an area of risk. Studies are lacking to show one that one treatment is more effective than the other, thus the decision is up to the surgeon. If an enteric closure (enterotomy, gastrotomy, anastomosis) is the source of primary leakage, the site should be debrided and revised as necessary. In some instances this may require a complete revision of an anastomosis.
If the source of infection cannot be resected, for instance a hepatic or prostatic abscess, the omentum can be used after debridement to both limit further contamination of the abdomen as well as bring blood supply and thus inflammatory support to clean up residual infection. The omentum should be secured with fine absorbable suture to the site to prevent inadvertent removal during patient recovery. If necessary, the omentum can be released from one side of its attachment to the greater curvature of the stomach to allow greater reach into the caudal abdomen. This can be particularly helpful if a partial omentectomy was necessary to remove contamination from the abdomen. The omentum is especially good at trapping gross debris and fibrin clots, but it can be all but impossible to ‘clean’ this material from the omentum necessitating resection of heavily contaminated areas. Leaving at least some omentum is desirable in septic peritonitis due to its surface area and absorptive capacity and ability to aid in sepsis resolution.
Adequate lavage is just as critical to survival in septic peritonitis patients as is definitive repair of the inciting cause. We cannot hope to eliminate all bacterial contamination, even if the peritonitis is acute. We can hope to reduce bacterial load and remove gross debris to the point where the inherent protective mechanisms of the abdomen can resolve the remainder. Warm sterile saline at a minimum of 200 ml/kg is the gold standard for lavage, or the surgeon can lavage until the recovered saline is clear. This is not only an opportunity to remove bacteria and gross contamination but also to warm the patient. Saline should be kept warm throughout surgery as tepid lavage fluid can cool the patient resulting increased risk of complications such as surgical site infection, prolonged recovery from anesthesia, hypotension and tissue ischemia and increased mortality. Moistened lap sponges can also be used if there is still significant gross contamination. A moist laparotomy sponges is gently moved through the abdomen, especially in the gutters and around the dorsal aspect of the liver lobes to collect fibrin rafts and gross debris that may have escaped suction. Care is taken to avoid abrading serosal surfaces with the lap sponge. A post-abdominal culture is collected prior to closure.
Septic abdomen patients can be closed primarily with or without suction drains, or they can be treated by open peritoneal drainage. This decision is sometimes made purely on financial restrictions in veterinary medicine as open peritoneal drainage requires significant nursing care and a second surgery for closure. In companion animals, open peritoneal drainage is generally reserved for those cases with severe systemic manifestations of peritonitis, ineffective drainage with conventional measures and extensive suppuration or necrosis of the deep layers of the abdominal incision. A patient managed as an ‘open abdomen’ is partially closed (loose non-absorbable suture in the linea allowing 1.5-2.5 cm gap between edges) and bandaged with sterile laparotomy sponges and a protective barrier layer. The bandage is changed under sedation and sterile conditions at least once daily and up to four times daily as drainage requires. Vacuum assisted closure can also be utilized for open abdomen management which allows fewer bandage changes but does not allow the clinician to visualize the abdomen at each bandage change and physically break down fibrinous adhesions. Definitive closure is performed when drainage improves both in volume and character, with the average duration of open abdomen management being 4 days in the veterinary literature. At the time of second surgery, a complete exploration is performed with lavage and suction and inspection of key sites prior to closure.
More commonly, dogs and cats with septic peritonitis are managed with primary closure and a closed suction drain system. This approach allows measurement of drain production, analysis of fluid and removal without the need for a second anesthesia or surgical procedure. Closed suction drains (such as Jackson Pratt style) are placed via separate incisions parallel to the midline incision and secured in place with suture to ensure an airtight and watertight seal. Fluid production is measured and recorded and compared on a ml/kg/day basis. Emptying the grenade and resetting the suction should be performed by experienced personnel using aseptic technique. There is no published rate of fluid production in closed suction drains in a septic peritonitis patient that indicates when to remove the drains. Instead, they are removed when the cytology of drain fluid (collected by suction from the line, not from the grenade) shows an absence of bacteria and non-degenerate neutrophils at the same time as a steady decrease in fluid volume. Drain production rarely reaches zero as the drain itself is a stimulus for inflammation and fluid production. A typical septic peritonitis case in our hospital has JP drains removed around day 3 or 4, but drains can remain useful up to two weeks. Closed suction drains can also be removed in stages if one drain is persistently producing very little fluid. At the time of removal, the site is cleaned in standard sterile fashion, the suture removed and the drain pulled using sterile technique. The tip of the drain is cut (with sterile scissors) into culture media for submission. After removal the drain site is stapled shut with a single skin staple. Primary closure with no drain system for septic peritonitis cases is reserved for those patients where the source of infection/contamination was removed and the degree of inflammation felt to be minimal, as with an acute perforation in an otherwise healthy animal, or when financial restrictions require minimal hospitalization following surgery. Due to the risk of iatrogenic injury, septic peritonitis cases are not sent home with closed suction drains in place.
Postoperative management of septic peritonitis patients can be a rollercoaster of pain management, nutritional requirements, hypoproteinemia, electrolyte derangements, transfusion medicine, hypotension and hypothermia and diagnostic result interpretation. These patients require intensive nursing care and constant adjusting of therapy that can absorb multiple ICU team members’ time and energy. Success in these challenging cases is a reflection of decision making and implementation from the moment the patient arrives through the complex balance of needs during the days following surgery.
In spite of significant advances in diagnostics and critical care management in the past few decades, survival rates for septic peritonitis in dogs and cats have not improved correspondingly. Reported mortality rates for dogs and cats range from 20% to 68% making a discussion with clients particularly difficult when pet owners ask for “numbers” on which to base their decision.
Attempts to analyze mortality rates across published retrospectives (meta-analysis or systematic review) have been unsuccessful in developing a single prognosis as inclusion criteria and level of care varies widely between studies and even within studies. There is reasonable evidence to suggest that prognosis should be considered differently for dogs and cats as well as for different origins of peritonitis (e.g. gastrointestinal vs. urogenital). Care decisions in post-operative management of septic peritonitis are frequently made based on the client’s financial limitations which further clouds true prognosis.