Supported by an unrestricted educational grant from Lederle, a division of Wyeth-Ayerst Laboratories

Volume III, Issue 1 - March 1996

Intraabdominal Infections-A Surgical Perspective

Intraabdominal infections are among the most difficult infections to diagnose early and treat effectively. A successful outcome depends on early diagnosis, rapid and appropriate surgical intervention, and selection of efficacious antibiotic regimens. Mortality rates associated with intra-abdominal infections range from 3.5% in patients with early infection following penetrating abdominal trauma to more than 60% in patients with well-established infection coupled with resultant multiple organ failure.(1)

These deep-seated infections generally occur after the continuity of the gastrointestinal (GI) tract is interrupted by trauma, intrinsic disease, or surgery (Table 1). The leakage of the endogenous microflora into adjacent tissues appears to overwhelm the host's defense mechanisms, resulting in infection. The degree of peritoneal infection dissemination depends primarily on five factors (Table 2). When dissemination is controlled and the initial event is promptly treated with appropriate surgical intervention and parenteral antibiotics, the chance for subsequent localized abscess decreases.

One of the most exciting and rewarding microbiological observations of the 20th century was the elucidation in the early 1970s of the role human anaerobic endogenous microflora play in infections, especially in intra-abdominal infections.(2)

Human Endogenous Gastrointestinal Microflora
The numbers and types of microorganisms increase progressively down the GI tract. In the healthy human the stomach and proximal small intestine support a rather sparse bacterial flora that includes both aerobes and anaerobes (< 10^4/mL).(3) Acidity and motility appear to be the major factors that inhibit the growth of bacteria in the stomach.

The gastric microflora are composed primarily of oral, penicillin-sensitive anaerobes and aerobic coliforms. The microflora of the distal small bowel represents a transitional zone between the microflora of the upper and lower GI tracts; modest numbers of aerobic and anaerobic microorganisms ( 10^3 to 10^8/mL) are usually detected. Concentrations of microorganisms are highest in the colon, where up to 10^8 aerobic coliforms and 10^11 anaerobes/g of stool or /mL of intestinal aspirate can be found.(3) The solid intraabdominal organs, including the liver and spleen, rarely harbor a significant population of endogenous microorganisms.

Infecting Flora of Intraabdominal Infections
In 1938 Altemeier became the first investigator to stress the polymicrobial aerobic and anaerobic nature of the bacterial flora of peritonitis resulting from acute appendiceal perforation.(4) These important studies were not elaborated on for nearly three decades, when modern techniques for isolating and growing anaerobic bacteria allowed better classification of these organisms in both normal flora and postoperative infection.

The number of aerobic and anaerobic bacteria isolated depends on the nature of the microflora of the diseased or traumatized organ. A complex polymicrobial flora results from GI tract contamination. The multiplicity of pathogens involved is evident from several reports (Table 3).

The aerobes commonly isolated in all of the summarized studies included E coli, klebsiella, streptococcus, proteus, and enterobacter species; the anaerobes most frequently isolated were bacteroides, peptostreptococcus, and clostridium species. Along with the other species of bacteroides, B fragilis--the single anaerobe most often isolated--accounted for 30% to 60% of all the anaerobe isolates in these studies. Purely anaerobic sepsis was reported in <15% of cases whereas purely aerobic infections were noted in <10% of the cases.

In addition, highly antibiotic-resistant strains of organisms such as Pseudomonas aeruginosa, Serratia marcescens, acinetobacter species, and providencia species are frequently isolated from patients who develop nosocomial sepsis. Persistent peritonitis in patients with long-term hospitalizations, repeated courses of antibiotics, multiple surgeries, and/or admissions to the intensive care unit favor the growth of breakthrough microorganisms such as Staphylococcus epidermidis as well as enterococcal and candidal species.

Clinical Features
Abdominal pain is present in all cases of peritonitis. When localized to one region of the abdomen, it is associated with rebound tenderness and muscle guarding in the area of organ perforation. In generalized peritonitis the pain and tenderness are found over the entire abdomen and are aggravated by any movement, including coughing. Associated muscle rigidity, in extreme cases, may result in the so-called boardlike abdomen. Patients with peritonitis also have a fever and progressive tachycardia that occurs secondary to third-space fluid losses within the peritoneal cavity.

Diagnosis
The diagnosis of diffuse peritonitis is usually based on the clinical history and physical findings. Associated increases of the peripheral leukocyte counts are frequently observed, most often exceeding 15,000 cells/mm(3) with a shift to the left. Basic radiographic examination may reveal free air below the diaphragm or locate mild intestinal distension associated with an intraperitoneal fluid collection between the bowel loops. Specialized radiographic procedures (ultrasound, CAT scan, and MRI) are most helpful in searching for localized intraabdominal collections and are rarely necessary to make the diagnosis of peritonitis.

Paracentesis, however, may be required to confirm the diagnosis. If gross pus, intestinal contents, or feces is aspirated the diagnosis is confirmed. A Gram stain of the aspirated fluid will offer immediate insights concerning the etiology of the peritonitis. Bacterial cultures should always be performed so further therapy can be based on knowledge of the pathogens and their antibiotic susceptibility.

Treatment of Intraabdominal Infection
The most important factors for treating intraabdominal sepsis are early diagnosis and prompt surgical intervention. The selection of the initial surgical procedure depends entirely on the nature of the pathology identified. The goals of the procedure should be to stop peritoneal contamination, debride necrotic tissue, remove debris and foreign bodies, and drain all localized collections of pus.

Antibiotic Selection
Unlike patients with superficial abscesses, in whom surgical drainage alone usually suffices, those with intraabdominal sepsis are best managed by a combination of surgical repair, diversion, and/or drainage in addition to the use of appropriate parenteral antibiotics. Treatment should be initiated as soon as the diagnosis is made (before surgery) and continued into the postoperative period. The ideal agent(s) and duration of therapy remain somewhat controversial. However, early experimental and subsequent clinical studies have indicated the spectrum of chosen antibiotics' activity must encompass both colonic aerobes and anaerobes, including B fragilis.

Rather than recount the results of the many studies reporting equal efficacy of various antibiotics in intraabdominal sepsis, we offer a list of the agents that are commonly used (Table 4).

In most recent studies seeking the ideal antibiotic for treating peritonitis, the abundant number of patients with penetrating abdominal trauma has served as the "battleground." While efficacy in this setting of acute contamination (usually involving otherwise healthy young men) does not necessarily translate into successful treatment of the severe infections that follow organ perforation in older, frequently immunodepressed patients, many lessons have been learned from these studies. One of the first was the importance of using antibiotics effective against both the aerobic and anaerobic bacteria of the GI microflora. A prospective study of 100 patients with abdominal trauma reported in 1973 revealed a significantly higher rate of anaerobic septicemia and intraabdominal infection among patients treated with cephalothin/kanamycin (aerobic coverage) than among those treated with clindamycin/kanamycin (aerobic-anaerobic coverage).(5) The authors appropriately stated, on the basis of these results, that anaerobic bacteria appeared to be a significant cause of infection in abdominal trauma. Few informative, well-controlled prospective data on this subject were cited in the surgical literature during the following decade.

Hofstetter and coinvestigators reported on their prospective study of 119 patients sustaining abdominal trauma.(6) The results obtained with this heterogeneous group indicated a short course of therapy with a single drug, cefoxitin, was as safe and effective as a course of a triple-drug regimen including an aminoglycoside, clindamycin, and ampicillin. The authors also suggested it might be prudent to consider leaving open the skin and subcutaneous tissues of patients who had hollow viscus injury because of the high incidence of localized wound infections.

Prospective randomized studies of penetrating abdominal trauma during the same period yielded similar conclusions regarding the efficacy of antibiotic therapy.(7) Differences in the efficacy of various cephalosporins were thought to be due to the varying activity of these agents against B fragilis. The antibiotic agents used in these studies generally lacked efficacy against enterococci. Despite the frequent isolation of these bacteria from infected sites, it was rarely necessary to alter the original antibiotic therapy for a successful outcome. In penetrating abdominal trauma or other situations in which there is contamination but not established infection, cultures appear unable to predict the causative pathogen(s) if an infection does occur. However, when there is established peritonitis, initial culture results are able to forecast response to antimicrobial therapy.

Finally, the choice of individual antibiotics or combinations must be influenced by many factors, including efficacy, toxicity, local nosocomial patterns of microbial sensitivity, and price.

Mechanical Surgical Techniques
Many mechanical techniques designed to reduce the bacterial burden within the peritoneal cavity have been advocated in addition to the primary surgical procedure (Table 5). The first of these techniques, intraoperative peritoneal irrigation with saline, antibiotic solutions, and/or povidone-iodine, is almost universally used to manage patients with secondary bacterial peritonitis.

The pendulum appears to be swinging back towards irrigation with crystalloid solutions alone. The irrigation, which takes place at the end of the operative procedure, involves either a pour-in technique or a jet-lavage using 2 to 3 liters of irrigant. Use of very large volumes of irrigation fluid ( > 5 L) has been thought to be associated with fluid and electrolyte shifts due to peritoneal absorption as well as the washing out of important peritoneal opsonins. The evidence does support reduction in the number of bacteria present, especially when irrigation is undertaken before fibrin is deposited in the peritoneal cavity.

Leiboff and Soroff reviewed 39 studies of closed postoperative peritoneal lavage for generalized peritonitis.(8) All of these studies used mortality as the primary variable for evaluating the technique. Twenty-seven studies were noncomparative, eight were nonrandomized and comparative, and four were prospective and randomized. The results indicated the better designed studies had poorer outcomes with closed postoperative lavage.

Radical surgical debridement was first advocated in 1975.(9) After initially suctioning free peritoneal fluid and pus and identifying and eliminating the source of contamination, tedious, time-consuming debridement of the entire peritoneal cavity was carried out successfully (all were cured) in 92 patients with generalized peritonitis. Systemic antibiotic treatment and peritoneal irrigation with physiologic saline were also used in each case. None of the infections recurred. To date, no clinical study has confirmed these remarkable results and one prospective randomized study has shown no advantages of the radical over the conservative approach.

Another technique recommended for acute generalized suppurative peritonitis was leaving the peritoneal cavity open.(10) At the conclusion of the surgical procedure, three or four layers of a 4-inch gauze pack are placed under the peritoneal aspect of the paramedian abdominal incision. Wires (3/0) are placed 1.5 to 2 cm apart through the peritoneum and abdominal wall except for the skin and subcutaneous fat. The wires are not tied initially but do hold the pack in position. External dressings, applied over the wound, are held in place with a binder. Peritoneal exudates drain freely for 48 to 72 hours, at which time the gauze pack is removed and the wires are tied to close the opening to the peritoneal cavity.

One of the first prospective studies of scheduled multiple laparotomies with abdominal lavage for the treatment of diffuse peritonitis was reported in 1986.(11) As with other open-abdomen techniques, the purposes were to ensure exclusion of the infected source, promote maximal elimination of toxic necrotic material, and permit prompt recognition of complications so immediate counteractive measures could be implemented. The average number of re-explorations was about four per patient and overall mortality in very high-risk populations was deemed acceptable. We personally favor this technique for high-risk patients with diffuse peritonitis, believing daily "unzipping" in the operating room should be performed until peritoneal sepsis is absent.

Successful percutaneous catheter drainage of intraabdominal abscesses, aided by computerized tomography or ultrasonography, has been reported since the early 1980s. As long as the abscess is accessible, this method appears to be an effective alternative to surgical incision and drainage.

It should be remembered that the primary purpose of these mechanical techniques is to lower the bacterial burden within the peritoneal cavity, thereby allowing parenteral antimicrobial agents to exert their maximal effect.

Summary
Intraabdominal sepsis most frequently follows abdominal trauma, perforated appendicitis, or diverticulitis. The initial leakage of the endogenous gastrointestinal microflora into the peritoneal cavity results in peritonitis and secondary septicemia, which is frequently followed by localized abscess. These infections are most frequently polymicrobial and relate directly to the unique endogenous microflora at the various levels of the GI tract. Treatment is primarily centered around prompt, appropriate surgical intervention. Parenterally administered antibiotics are also required to decrease the chance of local bacterial infection or septicemia. The choice of appropriate initial agent(s), before the results of culture and sensitivity reports are available, depends primarily on the clinical presentation and also on whether the infection occurred in the community or was nosocomial. Clinical and experimental studies have largely stressed the use of antibiotics with a spectrum of activity effective against the aerobic coliforms and anaerobic B fragilis.

Surgeons also employ various mechanical techniques to lessen morbidity and mortality of severe infections. Today, these most often include irrigation of the peritoneal cavity with varying amounts of solution at the time of the primary operation as well as the placement of a temporary "zipper" to provide easier access to the peritoneal cavity at subsequent operative procedures.

Ronald Lee Nichols, MD
William Henderson Professor of Surgery
Professor of Microbiology and Immunology
Tulane University School of Medicine
New Orleans, Louisiana

Intraabdominal Infections-A Surgical Perspective

Suggested Reading