Leave a comment



Anatomy and Physiology of the Peritoneal Space
Spontaneous (Primary) Bacterial Peritonitis in adults
Secondary Bacterial Peritonitis
Normal Gastrointestinal Flora
Antimicrobials in Secondary Bacterial Peritonitis
Surgical Management of Secondary Bacterial Peritonitis
Continuous Ambulatory Peritoneal Dialysis

Peritonitis is a commonly encountered illness that occurs in all age groups and is often infectious. Major types of peritonitis discussed in this chapter are spontaneous (primary) bacterial peritonitis (PBP), peritonitis complicating visceral perforation (“secondary” bacterial peritonitis), and that complicating continuous ambulatory peritoneal dialysis (CAPD). An appreciation of the anatomy, pathophysiology, clinical manifestations, and bacteriology of these diseases is important for appropriate management.
Anatomy and Physiology of the Peritoneal Space
The peritoneum is a closed space with many invaginations and outpockets. In females, the fallopian tubes breach this closure. Upper and lower peritoneal areas are connected by left and right gutters; these are potential conduits for infected material. The most dependent of these areas is the pelvis. Other candidates for accumulation of drainage are the left and right subdiaphragmatic spaces. The lesser sac, one of the largest of the potential spaces, is bounded by the pancreas and stomach and has an opening called the foramen of Winslow. Because of its unique location, it may be spared from general peritoneal infection. Alternatively, it may become infected as an isolated area.
The peritoneal cavity is lined by a single-layered serous membrane that allows rapid bidirectional transfer of materials. Physical forces, including oncotic and hydrostatic pressure, determine flow rate and direction. Antimicrobials such as quinolones, trimethoprim-sulfamethoxazole (TMP-SMX), aminoglycosides, b-lactams, clindamycin, and chloramphenicol penetrate the inflamed peritoneum and can rapidly achieve therapeutic concentrations. Thus, there may be no need for intraperitoneal antimicrobials in the management of some forms of peritonitis. The lymphatics remove proteins and particulate matter. Peritoneal lymphatics interdigitate with those above the diaphragm and allow the rapid dispersal of particulate matter into the pleurae. The diaphragmatic surface is covered with specialized lymphatics bearing stomata of 8 to 12 µm. Bacteria and proteins can be removed through pores of this size.
The peritoneal cavity responds to infection in several ways. Removal of potential pathogens occurs primarily by lymphatics. As an example, at least 50% of an intraperitoneal bacterial challenge is cleared into the bloodstream within 1 hour. Organisms are taken up through peritoneal lining cells, absorbed by lymphatics, and ultimately enter the bloodstream via the thoracic duct. Containment of infection is aided by production of fibrin secondary to inflammation. Normal anatomic barriers such as omentum, abdominal organs, and diaphragm may allow actively infected areas to remain sequestered from the remainder of the peritoneal cavity. Finally, host defenses that include peritoneal macrophages, polymorphonuclear leukocytes, complement, and immunoglobulins can be activated to opsonize, phagocytize, and kill microorganisms.
Spontaneous (Primary) Bacterial Peritonitis in adults
PBP is that unassociated with a primary intraabdominal source. Adults most often have cirrhosis with ascites as its substrate. In this population, it is a marker of severe liver disease. In-hospital mortality reaches 50%, and rates of relapse of up to 43% at 6 months and 69% at 1 year and 1-year mortality rates of approximately 60% are published. Patients at particular risk are those with ascitic protein concentrations of less than 1 g/dL. Patients with severe liver disease who survive an initial bout of PBP may be candidates for liver transplantation. Symptoms include worsening hepatic failure, abdominal pain or tenderness, and fever. These classic findings may be minimal, however, and the disease occasionally may be diagnosed in the absence of symptoms. Generally, paracentesis should be performed in patients with ascites and any of the classic findings.
PBP most commonly is caused by a single microbe. Identification of multiple enteric bacteria should prompt a search for perforated viscus. Escherichia coli (40% to 60%) and streptococci/enterococci (30%) are most often seen. Other enteric gram-negative bacilli, especially Klebsiella species, comprise most of the remainder. Anaerobes are rarely identified, probably because of high oxygen tensions in ascitic fluid.
A variant of spontaneous peritonitis, culture-negative neutrocytic ascites, is defined as the combination of ascitic fluid leukocytosis (>500/mm3), no prior antibiotics, and negative cultures. Presentation and natural history are the same as with PBP. The reason for negative cultures involves the immune response of ascitic fluid.
Blood cultures should be performed for patients suspected of having PBP, and individuals should undergo diagnostic paracentesis. The former are positive in about a third of patients. Within ascitic fluid, cell count and differential demonstrate primarily polymorphonuclear leukocytes and cell counts of more than 250 to 500/mm3. Marked peripheral leukocytosis does not affect this. Other parameters that help discriminate between PBP and other ascitic syndromes include an ascitic fluid-serum lactate dehydrogenase ratio above 0.4 and an ascitic fluid-serum glucose ratio below 1.0. Generally, PBP by these criteria has been exudative. Gram’s stain typically demonstrates a single morphology (implying a single organism). Paracentesis fluid should be routinely injected into blood culture bottles, and especially the BacT/ALERT system. This results in faster and greater yields than does the use of traditional agar inoculation systems.
Therapy consists of appropriate antimicrobials and supportive care. Choice of agent can be based on Gram’s stain. Gram-positive cocci in chains represent streptococci. Two to three million units of aqueous penicillin G IV every 4 hours, plus 5 mg of gentamicin per kilogram IV daily in a single dose provides optimal coverage until identification is completed. Streptococcus pneumoniae, S. pyogenes, and most other streptococci can then be treated with penicillin as monotherapy. Enterococcus faecalis or E. faecium generally requires combination therapy. Unfortunately, some strains of enterococci now produce b-lactamase or may be vancomycin-resistant. Therapy must be individualized based on susceptibilities. Gram-negative bacilli demonstrated on Gram’s stain can generally be managed with monotherapy. Options include third-generation cephalosporins, antipseudomonal penicillins, aztreonam, quinolones, TMP-SMX, ticarcillin-clavulanate, piperacillin-tazobactam, or imipenemcilastatin. Choice depends primarily on susceptibility patterns, cost, ease of administration, and patient factors.
Recommendations for empiric therapy if treatment cannot be guided by Gram’s stain are generally a third-generation cephalosporin or a b/b-lactamase combination. Aztreonam has been well studied for PBP, and although it works well for illness caused by gram-negative bacilli, gram-positive superinfection has been noted. Thus, its use as empiric monotherapy cannot be supported. For susceptible enteric gram-negative bacilli in patients tolerant of oral therapy, oral quinolones or TMP-SMX could be considered.
Duration of therapy is best guided by the results of a second paracentesis performed 48 hours after initiation of antibiotics. At this time, patients should have ascitic fluid WBC counts of less than 250/mm3 and negative cultures. If counts have risen or if cultures remain positive, a ruptured viscus should be considered. For patients with ascitic fluid counts of less than 250/mm3 at 48 hours, duration of treatment should be 5 days. Evaluation for ruptured viscus, resistant pathogens, and other reasons for failure should be performed when counts have not fallen at repeated (48-hour) paracentesis.
Primary peritonitis caused by Mycobacterium tuberculosis or fungi needs to be considered in patients with negative bacterial cultures and exudative ascites. Tuberculous peritonitis comprises more than 50% of cases of abdominal tuberculosis, is most common in women (71%), and can mimic an acute abdomen, tumor, or cirrhosis. Many of these patients have no evidence of disease elsewhere (46%), and skin tests may be negative (17%). Most cases are associated with exudative ascites and lymphocyte counts above 500/mm3. Diagnosis is best made by peritoneal biopsy for smear, culture, and histopathology. Fungal peritonitis is most commonly associated with either Candida species or Cryptococcus neoformans. The former should be considered in patients with previous intraabdominal surgery and recent broad-spectrum antimicrobial treatment. The latter usually presents a component of disseminated cryptococcosis and is generally diagnosed by aspiration of ascites. Severe underlying hepatic disease is noted, and there may be a history of recent upper gastrointestinal bleeding.
Prevention of PBP should be attempted in all patients who survive an initial bout or who are high risk (variceal bleeding, low ascitic protein concentrations, or prolonged prothrombin times). Oral quinolones and TMP-SMX have been utilized. The latter, at a dose of 1 DS 5 d/wk was associated with a statistically lower likelihood of PBP than placebo, but long-term mortality rates were unchanged. Diuresis to decrease ascitic fluid volume (and therefore raise ascitic fluid protein) is also indicated.
Secondary Bacterial Peritonitis
Secondary bacterial peritonitis generally results from spillage of the contents of a hollow viscus. Common causes are penetrating trauma, malignancy, diverticular and appendiceal infections, cholecystitis, and pyloric ulcer disease. Consequences depend in part on the bacteriologic composition of the spilled material. Clinical presentation often involves well-defined complexes of symptoms. The elderly and those on high doses of corticosteroids may present in more subtle fashions. In these populations, abdominal pain may be blunted and fever may be lower. As a result, symptoms may last longer before diagnosis, and patients may therefore be sicker on presentation.
Normal Gastrointestinal Flora
Organisms colonizing the gastrointestinal tract vary qualitatively and quantitatively among different sites (Table 16-1). Such differences have important clinical implications. Polymicrobial contamination with three to five species often follows large-bowel spillage. Anaerobes, including Bacteroides fragilis, predominate. Pseudomonas aeruginosa may be an important consideration in patients with complicated appendicitis.

Table 16-1. Comparative bacteriology of the intact gastrointestinal tract

Summarized results from animal and human trials indicate the following: (a) Early mortality from peritonitis approaches 40% and is caused primarily by gram-negative sepsis from E. coli; (b) intraabdominal abscesses develop in most survivors with a complex flora that includes B. fragilis, E. coli, enterococci, and other anaerobes; (c) the capsule of B. fragilis is a virulence factor associated with abscess formation; and (d) there exists a pecking order of antimicrobial agents that vary considerably in their ability to improve survival and decrease abscess formation among survivors and to decrease the number of viable bacteria within experimental abscesses. Conclusions from these studies demonstrate the need for antimicrobials that target both enteric gram-negative and anaerobic (including B. fragilis) components of these infections.
In community-acquired peritonitis, the roles of enterococci and Candida are controversial. Enterococci are identified in up to 20% of infections, and some recent data identify documentation of enterococcal species as a risk factor for adverse outcome when they are not targeted in therapy. However, there are no data demonstrating enhanced outcomes when enterococcal infection is treated. In general, neither organism needs to be covered empirically unless it is believed to be a predominant pathogen on Gram’s stain, associated with positive blood cultures, or noted in pure culture. Vancomycin-resistant enterococci are unlikely to be identified in community-acquired disease. However, both organisms become important considerations following antimicrobial therapy or reoperation for complications.
Antimicrobials in Secondary Bacterial Peritonitis
Antibiotics employed in secondary bacterial peritonitis should take into account which pathogens are likely; this is based in part on the initial site of infection. As an example, perforation of a previously healthy stomach is likely to result in either sterile peritonitis or infection associated with low numbers of oropharyngeal flora. Alternatively, peritoneal contamination from the colon will be associated with a complex flora involving enteric bacilli and anaerobes. Numerous antibiotics (when coupled with drainage), administered as monotherapy or in combination, are effective and appropriate for the management of secondary bacterial peritonitis. The author prefers monotherapy when possible, as it is generally easier to manage and may be less expensive. Table 16-2 summarizes some initial treatment regimens. Choice among them requires knowledge of local resistance problems, host factors (allergy, end-organ function), costs, and availability in hospital formularies.

Table 16-2. Antimicrobials useful in secondary community-acquired bacterial peritonitis

Optimal length of therapy for secondary bacterial peritonitis is unknown. Current data suggest that 5 to 7 days of IV therapy is sufficient for infection associated with recent penetrating trauma or peritonitis in otherwise healthy persons. Complicated peritonitis with residual abscess formation needs longer treatment. Use of oral agents for part of the course may be sufficient in patients with functional gastrointestinal tracts.
Surgical Management of Secondary Bacterial Peritonitis
Removal of necrotic tissue, drainage of abscesses, and closure of perforations are major goals of surgery. Penetrating trauma or acute perforation of a viscus requires exploratory laparotomy. Antimicrobials are administered as soon as possible and can often be discontinued by 5 days.
Percutaneous catheter drainage of abscesses should be employed when technically feasible, but this technique has little value in acute secondary peritonitis. For drainage of intraabdominal abscesses, success rates in noncomparative studies are more than 85%. Percutaneous drainage should be strongly considered as initial therapy for patients with approachable intraabdominal abscess and for those who are poor surgical candidates.
Continuous Ambulatory Peritoneal Dialysis
CAPD-related peritonitis is a significant complicating factor of this procedure and is noted approximately 1.7 times per patient-year or about every 7 to 10 months. Approximately 60% of patients will have a bout of peritonitis during the first year of CAPD. Recurrences develop in 20% to 30% of individuals. The relationship between CAPD peritonitis and mortality is most noted in white, nondiabetic, older patients. Some data suggest that peritonitis contributes to mortality in about 15% of patients and is most notable with gram-negative bacilli and fungi. In selected patients on long-term CAPD, peritonitis does not develop, presuming a role for host defenses or meticulous care. Most common pathogenetic mechanisms are migration along the dialysis catheter or breaks in sterile technique during dialysis exchanges. Dialysis fluid inside the peritoneal cavity can support the growth of many pathogens, including most enteric bacilli, Staphylococcus aureus, and P. aeruginosa. Patients identified as nasal carriers of S. aureus are at higher risk than noncarriers for the development of S. aureus peritonitis. Data also suggest that peritonitis is more likely to develop in patients with enhanced anxiety and decreased quality-of-life scores. The reasons for this are uncertain.
In the context of CAPD, peritonitis can be defined as the presence of turbid dialysate when etiologies other than infection cannot be identified; this condition occurs in the presence of WBC counts of more than 300/mm3. This definition does not require the presence of constitutional symptomatology, leukocytosis, abdominal pain, or a positive Gram’s stain or culture. Risk factors include advanced age, use of CAPD rather than continuous cycling peritoneal dialysis (CCPD), and earlier initiation of CAPD in the patient’s history. The reasons why peritonitis never develops in up to 50% of patients while others have recurrent episodes are uncertain; however, chronic nasal carriage of S. aureus has been identified as a risk factor in some patients with recurrent infections associated with this organism.
Clinical presentation is generally cloudy peritoneal dialysate. Fever is noted in 33% of patients, and abdominal pain and tenderness are seen in the majority. Up to one third of patients will be sick enough to require hospitalization, but therapy is generally rendered in an outpatient setting. It is uncommon for infections to disseminate beyond the peritoneal cavity.
The bacteriology of peritonitis complicating CAPD is generally monomicrobial. The presence of polymicrobial infection should prompt an assessment for perforation of a viscus. S. aureus and Staphylococcus epidermidis are most commonly implicated and are responsible for approximately 50% of infections. Up to 70% of isolates are gram-positive bacteria. P. aeruginosa is associated with 5% to 10% of cases but causes significant mortality and morbidity. Miscellaneous organisms, including fungi (mostly Candida species) and mycobacteria (not M. tuberculosis), and sterile specimens are noted in 10% to 20% of cases. Fungal peritonitis represents 3% to 4% of all cases of peritonitis, is mostly caused by Candida species, and is associated with prior use of antibiotics. Removal of catheter, short-term hemodialysis, and catheter replacement at 2 to 8 weeks appears to render satisfactory therapy. Etiology can usually be suspected from Gram’s-stained specimens of fluid. This information should be utilized for antibiotic decision making. Culture for aerobic and anaerobic organisms should always be obtained, and special cultures for acid-fast bacilli and fungi should be sought if standard Gram’s stain fails to demonstrate bacteria.
The linchpin of management is the administration of appropriate antimicrobials in doses sufficient to exceed the minimum inhibitory concentration for the offending organism at the site of infection for at least part of the dosing interval. Most cases unlikely to be complicated by bacteremia can be managed by intraperitoneal antimicrobials and preservation of the dialysis catheter. P. aeruginosa infections are generally managed with two effective agents plus catheter removal.
Table 16-3 summarizes the treatment recommendations for selected antibiotics. Empiric therapy with cefazolin and gentamicin is sensible for patients without positive Gram’s stains. Major indications for catheter removal include infection caused by fungi, mycobacteria, P. aeruginosa, or Corynebacteria. Additional reasons are perforated viscus, relapse of peritonitis, and tunnel infection. Parenteral antimicrobials are initially indicated if bacteremia or sepsis is suspected. Duration of treatment may be 7 to 10 days (gram-positive infection) or 2 to 3 weeks (gram-negative or fungal infection). Amphotericin B in a total dose of up to 500 mg following catheter removal remains the gold standard for susceptible fungal infections.

Table 16-3. Antibiotics for use in peritonitis complicating continuous ambulatory peritoneal dialysis

Prevention of peritonitis in CAPD is difficult and best associated with education and adherence to appropriate techniques. Antibiotic prophylaxis does not work well and is generally not recommended. In patients with S. aureus peritonitis, eradication of identified nasal carriage should be attempted with mupirocin. S. aureus peritonitis may also be preventable with use of daily mupirocin at the catheter exit site or by use of 600 mg of rifampin for 5 days of every 3 months. (R.B.B.)
Antillon MR, Runyon BA. Effect of marked peripheral leukocytosis on the leukocyte count in ascites. Arch Intern Med 1991;151:509–510.
This investigation focuses on 29 patients who underwent paracentesis when peripheral leukocyte counts were higher than 20,000/mm3. None had peritonitis as a cause of ascites, and leukocyte counts in ascitic fluid remained low despite the peripheral findings.
Bhuva M, Ganger D, Jensen D. Spontaneous bacterial peritonitis: an update on evaluation, management, and prevention. Am J Med 1994;97;169–175.
The authors present an excellent overview of spontaneous bacterial peritonitis. Patients with cirrhosis and ascites with fever or abdominal pain should generally undergo paracentesis. Prophylaxis with antibiotics and diuresis may decrease mortality. When possible, drugs without nephotoxic potential should be employed for prophylaxis and treatment.
Cooper GS, Shlaes DM, Salata RA. Intraabdominal infection: differences in presentation and outcome between younger patients and the elderly. Clin Infect Dis 1994; 19:146–148.
The authors conducted a retrospective study of about 130 eligible patients discharged from a tertiary care center with intraabdominal infections and compared presentation, bacteriology, antibiotic use, and outcomes between younger patients and those over age 65 (“elderly”). Elderly patients had a longer interval until therapy and presented with less obvious symptoms and signs of intraabdominal infection. Bacteriology and use of antibiotics was similar, and mortality rates were not significantly different. However, those over age 65 had prolonged hospitalizations, and normalization of their temperature took longer. Physicians should be aware of the fact that many serious bacterial infections present subtly in the elderly, and a high index of suspicion is warranted.
Goldie SJ, et al. Fungal peritonitis in a large chronic peritoneal dialysis population: a report of 55 episodes. Am J Kidney Dis 1996;28:86–91.
A retrospective review of records demonstrated that approximately 3% of cases of CAPD-related peritonitis were fungal. Most were caused by Candida species. Those with fungal peritonitis had more infections per year and were more likely to have received prior antibiotics. No other risks were identified. Catheter removal and antifungal treatment were mainstays of therapy, and catheter replacement after 2 to 8 weeks of temporary hemodialysis resulted in good salvage.
Inadomi J, Sonnenberg A. Cost-analysis of prophylactic antibiotics in spontaneous bacterial peritonitis. Gastroenterology 1997;113:1289–1294.
The authors conclude that prophylactic antibiotics are a cost effective strategy for patients with cirrhosis and ascites. Issues addressed include development of PBP, subsequent mortality, and the cost of drugs. Agents primarily assessed were norfloxacin and TMP-SMX. Prophylaxis is cost effective when compared with placebo.
Jakubowski A, Elwood RK, Enarson DA. Clinical features of abdominal tuberculosis J. Infect Dis 1988;158:687–693.
Eighty-one cases of intraabdominal tuberculosis were reviewed. Most were peritonitis. This was most commonly a disease of women, often not associated with disease elsewhere. More than 80% had positive tuberculin test results. Diagnosis is often difficult, and a long differential diagnosis is considered. Treatment is usually curative.
Johnson CC, Baldessarre J, Levison ME. Peritonitis: update on pathophysiology, clinical manifestations, and management. Clin Infect Dis 1997;24:1035–1047.
A contemporary review of spontaneous, secondary, and CAPD-associated peritonitis. A good review of issues covered in the title.
Keane WF, et al. Peritoneal dialysis-related peritonitis treatment recommendations: 1996 update. Perit Dial Int 1996;16:557–573.
The authors provide an exhaustive profile of complications and management strategies of peritoneal dialysis. Antibiotics and dosing regimens are provided, and doses are divided into both continuous and intermittent options. The authors recommend use of a first-generation cephalosporin plus an aminoglycoside as empiric therapy in most instances. They should be praised for recognizing the potential severe problems with vancomycin-resistant enterococci! They also address the general inadvisability of prophylactic antibiotics to prevent peritonitis in this population. An excellent single source of information about the subject.
McClean KL, Sheehan GJ, Harding GKM. Intraabdominal infection: a review. Clin Infect Dis 1994;19:100–116.
An excellent review of basic pathophysiology, bacteriology, diagnosis, and management of intraabdominal infections. Issues related to imaging are explored, and the authors devote significant space to surgical issues, including type of operation. Controversial issues such as the roles of enterococci and Candida species are raised, but no answers are provided. A segment on infections in immunocompromised patients is particularly valuable as an overview of pathogens in difficult hosts.
Runyon BA, et al. The serum-ascites albumin gradient is superior to the exudate-transudate concept in the differential diagnosis of ascites. Ann Intern Med 1992; 117:215–220.
Although this investigation does not specifically address peritonitis, the article is important because it describes a newer method to evaluate ascites. Assessment of the serum-ascites albumin gradient (high in portal hypertension) may be a better method to manage ascites. Part of the reason for this, the authors claim, is that many cases of bacterial peritonitis may not be associated with “exudative” ascites by classic criteria.
Silvain C. Can septicemia and ascitic fluid infections in cirrhotic patients be treated by the oral route alone? Gastroenterol Clin Biol 1989;13:335–339.
The investigator utilized oral antimicrobials, primarily quinolones, for the management of primary peritonitis. Outcomes were good.
Singh N, et al. Trimethoprim-sulfamethoxazole for the prevention of spontaneous bacterial peritonitis in cirrhosis: a randomized trial. Ann Intern Med 1995;122:595–598.
Patients with cirrhosis and ascites, stratified by additional risks, were enrolled in a prospective study to assess whether the use of prophylactic TMP-SMX resulted in fewer cases of PBP. Mean duration of follow-up was 90 days, and PBP or spontaneous bacteremia developed in significantly more patients without antibiotic (27% vs. 3%). Mortality rates were 20% versus 7%. This investigation suggests that antibiotic prophylaxis prevents adverse outcomes in high-risk patients. It should probably be considered a standard of care in this population.
Wilcox CM, Dismukes WE. Spontaneous bacterial peritonitis. A review of pathogenesis, diagnosis and treatment. Medicine (Baltimore) 1987;66:447–456.
One of the best reviews on all aspects of spontaneous bacterial peritonitis.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: