ACP medicine, 3rd Edition

Infectious Disease

Peritonitis and Intra-Abdominal Abscesses

  1. Conrad Liles M.D., Ph.D.1
  2. Patchen Dellinger M.D.2

1Associate Professor of Medicine, University of Washington School of Medicine

2Professor and Vice Chairman, Department of Surgery, University of Washington School of Medicine, Chief and Associate Medical Director, Department of Surgery, University of Washington Medical Center

The authors have no commercial relationships with manufacturers of products or providers of services discussed in this chapter.

March 2004


Peritonitis is a diffuse or localized inflammatory process affecting the peritoneal lining.1,2 Peritonitis has acute and chronic forms and may have a variety of causes, including pyogenic bacteria (e.g., Escherichia coli), tuberculosis, fungi, parasites (such as those from ruptured hepatic amebic abscesses or hydatid cysts), carcinomatosis, chemical irritation (such as that from bile), drug-hypersensitivity reaction, foreign-body reaction (such as starch peritonitis), and certain systemic illnesses (such as familial Mediterranean fever and systemic lupus erythematosus). Only acute peritonitis caused by bacteria or fungi, including primary and secondary peritonitis, is discussed here. Primary peritonitis has no underlying intra-abdominal disorder. Secondary peritonitis has an intra-abdominal focus that initiates the infection. Tuberculous peritonitis is considered elsewhere [see 7:II Infections Due to Mycobacteria].


Epidemiology and Etiology

Spontaneous bacterial peritonitis (SBP) is being recognized with increasing frequency in patients with either advanced chronic liver disease and concomitant ascites or fulminant hepatic failure and ascites.3,4 The underlying chronic liver disease is usually alcoholic cirrhosis in older patients, whereas postnecrotic cirrhosis predominates in children and young adults. SBP is both common and serious, occurring in 10% to 30% of patients with cirrhosis and ascites, with an associated mortality between 30% and 50%.2,5,6,7 The risk of SBP is increased in cirrhotic individuals with low ascitic protein levels (1 g/dl or lower).8,9 E. coli is the most common cause of SBP and is isolated in about half of patients. Pneumococcal and streptococcal species are each responsible for 15% to 20% of cases, Klebsiella species for about 10%, and anaerobic or microaerophilic organisms for about 5%. Staphylococcus aureus is an infrequent cause of SBP but a major cause of peritonitis in cirrhotic patients with LeVeen peritoneovenous shunts. A variety of other organisms, including Listeria monocytogenes, Campylobacter coli, and Aeromonas species, have been responsible for isolated cases of SBP. In most instances in which the causative organism is aerobic, a single organism is involved, and concurrent bacteremia is frequent.

Although primary peritonitis occurs most often in children, it can develop in adults; nearly all adult patients are women.10 Although many patients have had nephrosis, most have not had preexisting ascites. The source of infection is usually occult but may involve the female genital tract. The infecting organisms are almost always pneumococci or group A β-hemolytic streptococci; gram-negative bacilli are only rarely implicated. For reasons that are unclear, the incidence of primary peritonitis has decreased strikingly in the past several decades.

Occasionally, SBP develops in patients with systemic lupus erythematosus and lupus nephritis without detectable ascites, most of whom have been receiving corticosteroid therapy. The most common etiologic agents in this form of SBP are gram-positive cocci such asStreptococcus pneumoniae and group B streptococci. The most likely route of infection is bacteremic seeding of ascitic fluid, which may be precipitated by portal hypertension; intrahepatic shunting; intestinal bacterial overgrowth; and impaired host defense mechanisms, including diminished bactericidal activity in ascitic fluid.3,6 Less often, SBP results from transmural migration of enteric bacteria (possibly associated with diarrhea, a common symptom in cirrhosis). Severe liver disease, hepatocellular carcinoma, gastrointestinal bleeding, and a focus of infection in the urinary tract or elsewhere in the body increase the risk of SBP.9,11,12 Prior paracentesis may be contributory in a few instances. Penetrating lesions of the biliary tract, peptic ulcer disease, and overt bowel inflammation (such as appendicitis or diverticulitis) do not appear to be sources of infection.


Clinical features

The clinical presentation of SBP is often subtle.3,6,7 Although ascites is always present, the volume of fluid may occasionally be small enough to necessitate ultrasonography for confirmation. Fever is the most common symptom but is absent in more than 30% of cases. Abdominal pain and hepatic encephalopathy are present in most patients. However, only half of the patients with SBP have abdominal tenderness, and as many as one third may be free of signs and symptoms of infection. Hence, SBP should be suspected in any cirrhotic patient who presents with unexplained clinical deterioration or hypotension.

Laboratory tests

The key to diagnosis is examination of the ascitic fluid for bacteria and white blood cells. The polymorphonuclear leukocyte (PMN) count of the ascitic fluid is the best indicator of SBP. Although counts of more than 500 cells/mm3 are considered specific for SBP, counts of 250 cells/mm3 or more suggest a diagnosis of SBP and are considered specific enough to mandate treatment of SBP in cirrhotic patients in whom no other evidence of infection is present. An ascitic fluid PMN count below 250 cells/mm3 excludes the diagnosis of SBP.6,7

Because bacterial counts are often very low, Gram stain of ascitic fluid in SBP is typically negative. However, a Gram stain is always useful because visualization of a single bacterial type would be consistent with SBP, whereas the presence of multiple bacterial forms would suggest secondary peritonitis. Because of the low concentration of bacteria, cultures are best performed by inoculation of 10 to 20 ml of ascitic fluid into a blood culture or BACTEC bottle at the bedside.

Three variants of SBP have been recognized on the basis of ascitic fluid PMN counts and cultures. In typical SBP, the PMN count is 250 cells/mm3 or higher and cultures are positive. When the PMN count is 500 cells/mm3 or higher but cultures are negative, the syndrome is called culture-negative neutrophilic ascites (CNNA). When the PMN count is below 250 cells/mm3 but cultures are positive, the syndrome is termed bacterascites (BA). The clinical features and prognosis of SBP and CNNA are indistinguishable, and the two variants should be managed identically. In contrast, BA can be self-limited; if patients are asymptomatic, they can be managed with careful observation and repeat paracentesis after 48 hours. Antibiotic therapy can be initiated if clinical symptoms develop or if the PMN count of the ascitic fluid rises.

It is important to distinguish SBP from secondary peritonitis resulting from intra-abdominal disease, such as a perforated viscus. An ascitic-fluid white blood cell count of 10,000/mm3 or higher suggests secondary peritonitis, as does the presence of multiple bacterial species, anaerobes, or fungi. Patients should, of course, always be evaluated clinically and radiologically to exclude an underlying intra-abdominal process that might give rise to secondary peritonitis. Peripheral blood leukocytosis and positive blood cultures are common in both SBP and secondary peritonitis.

Differential Diagnosis

Bacterial peritonitis may be closely mimicked by acute pancreatitis, particularly in a patient with cirrhosis [see also Pancreatic Infections,below]. Abdominal pain, fever, rebound tenderness, hypotension, and peripheral leukocytosis are common in both bacterial peritonitis and acute pancreatitis. In a patient with pancreatitis, a diagnostic abdominal aspiration may even reveal cloudy fluid, but the turbidity is caused by floating fat globules derived from fat necrosis. Very high serum amylase levels are present in acute pancreatitis, but elevated levels also occur in peritonitis after intestinal perforation or obstruction and in the presence of renal failure.

In a patient with cirrhosis and ascites, a number of conditions may be mistaken for peritonitis, including acute peptic ulcer, cholecystitis, mesenteric artery occlusion, and other intra-abdominal processes. Paracentesis is helpful in arriving at a diagnosis in these circumstances.

Acute bacterial peritonitis may be distinguished from tuberculous peritonitis by several features. Tuberculous peritonitis is marked by a more indolent course, the absence of a peripheral leukocytosis, radiologic evidence of pulmonary tuberculosis, and a mononuclear response in the peritoneal fluid. In the patient with tuberculous peritonitis who does not have cirrhosis and ascites, the abdomen may have the characteristic so-called doughy consistency.

Peritonitis may be superficially suggested by the abdominal pain of acute porphyria, by lead colic, by diabetic acidosis, and by tabetic crisis, but the other features of these illnesses serve to distinguish them from peritonitis. The signs and symptoms of familial Mediterranean fever—high temperature, abdominal pain, abdominal guarding, and peripheral leukocytosis—may suggest bacterial peritonitis. The periodicity of familial Mediterranean fever and its occurrence predominantly in persons of Sephardic, Armenian, and Arab ancestry are helpful in differentiating it from bacterial peritonitis.

SBP may be difficult to diagnose in a patient with systemic lupus erythematosus who experiences acute abdominal pain and fever. These symptoms may stem from a variety of independent surgical problems (e.g., perforated ulcer, intestinal obstruction, and mesenteric occlusion) that must be distinguished from abdominal problems directly related to lupus, such as vasculitis, pancreatitis secondary to vasculitis or corticosteroid therapy, and SBP. Examination of peritoneal fluid obtained by paracentesis, by culdocentesis, or during laparotomy may be the only way to determine the presence of bacterial peritonitis.


Until culture results are available, broad coverage should be directed against enteric organisms. Nephrotoxic drugs, including aminoglycosides, should be avoided whenever possible.13 Cefotaxime (2 g I.V. every 8 hours) has emerged as the favored agent for the empirical treatment of SBP; alternative useful agents include ceftriaxone, ceftazidime, cefonicid, ceftizoxime, ampicillin-sulbactam, meropenem, and imipenem-cilastatin, as well as fluoroquinolones (i.e., ciprofloxacin, levofloxacin, gatifloxacin, and moxifloxacin).6,7Traditionally, intravenous antibiotics have been administered for 10 to 14 days, but 5 days of therapy appears to be as effective, provided that the patient is doing well clinically and that the ascitic fluid is sterile, with a PMN count that is below 250 cells/mm3 before discontinuance of antibiotics.6,7,14

Renal failure is a frequent complication of SBP; peripheral vasodilation and renal vasoconstriction are probably responsible.15 Infusions of albumin (1.5 g/kg at the time of diagnosis and 1 g/kg on day 3) can substantially reduce the incidence of renal failure and mortality in patients with SBP.16


Because patients with cirrhosis are at high risk for primary SBP and because recurrences develop in 43% of these patients within 6 months and in 69% within 1 year after an initial episode of SBP, both primary and secondary prophylaxis regimens are now recommended for certain subgroups of patients. In nonbleeding cirrhotic patients with persistent ascites after an initial episode of SBP, continuous secondary prophylaxis with oral norfloxacin (400 mg daily) is currently recommended. Alternative oral antimicrobial agents for prophylaxis include ciprofloxacin, levofloxacin, trimethoprim-sulfamethoxazole, and amoxicillin-clavulanate. In cirrhotic patients with upper gastrointestinal hemorrhage, primary prophylaxis with oral norfloxacin (400 mg every 12 hours) or alternative systemic therapy (ciprofloxacin, levofloxacin, ampicillin-sulbactam) for a minimum of 7 days is advised. Primary prophylaxis with norfloxacin or another fluoroquinolone should also be considered in cirrhotic patients with low ascitic protein levels (i.e., less than 1.5 g/L).6,7,17,18,19 Appropriate prophylaxis not only reduces the incidence of SBP but also improves overall survival.20



Secondary peritonitis occurs as a complication of intra-abdominal disease. It may result from appendicitis, diverticulitis, penetrating abdominal wounds, blunt trauma to the abdomen, perforation of the gastrointestinal tract (e.g., by a peptic ulcer or bowel neoplasm), or rupture of an intra-abdominal abscess. Secondary peritonitis can be divided into spontaneous cases caused by an underlying disease such as appendicitis or diverticulitis and cases that result from a ruptured viscus incurred from an operation, a procedure, or an episode of trauma.21 Most of these infections are polymicrobial. The pathogens include both anaerobic species (principally Bacteroides fragilis, peptococci, and peptostreptococci) and aerobic species (E. coli, Proteus species, Klebsiella species, and various streptococci and enterococci).22 Bacteremia—which occurs in only 20% to 30% of cases—is most commonly caused by E. coli, Bacteroides species, or both.23,24,25 The prognosis of secondary peritonitis depends on the underlying cause and the patient's physiologic response to the infection. Mortality is lowest in patients with appendicitis or perforated duodenal ulcer (10%) and highest in those with other intra-abdominal processes (50%) or postoperative peritonitis (60%). Both mortality and the likelihood of complications, including the necessity for a second operation, increase as the patient's physiologic response to the disease is more marked. This is most easily assessed using the Acute Physiology and Chronic Health Evaluation II (APACHE II) score.23,24

Tertiary peritonitis is a relatively new term that refers to the persistence of intra-abdominal infection after the initial surgical and medical treatment of secondary peritonitis.26 Not all authors discussing tertiary peritonitis agree on the definition of this syndrome. In a typical case of tertiary peritonitis, operative exploration in a patient with signs and symptoms of peritonitis after prior treatment for peritonitis will reveal inflammation and bacterial growth despite the absence of a focus for continuing infection, such as a perforated viscus, gangrenous tissue, or abscess. The organisms recovered tend to be considered nonpathogens and tend not to be typical of enteric flora. Tertiary peritonitis can be considered evidence for a type of host defense failure.


The clinical features of secondary peritonitis most commonly include peritoneal signs such as involuntary guarding, referred percussion tenderness, and abdominal tenderness. There may be abdominal distention. Testing for rebound tenderness is painful to the patient and unhelpful in diagnosis. The abdomen is often not silent. Fever and leukocytosis are usually present. Free air may or may not be visible on plain abdominal films. An ultrasound or CT scan showing peritoneal free fluid or gas in association with a compatible clinical picture confirms the diagnosis. If the diagnosis is clinically obvious, radiographic studies are not required.

In the absence of ascites, the peritoneum has a remarkable capacity to wall off and localize infection to a portion of the abdomen. Thus, in situations such as rupture of the appendix, a localized peritonitis develops and results in a periappendiceal or pelvic abscess [see Intra-abdominal Abscesses, below].


Peritonitis secondary to bowel perforation, ruptured gangrenous appendix, or penetrating trauma requires prompt surgical intervention in addition to antimicrobial therapy. Surgical therapy, known as source control, should be directed at correction of the underlying disease, debridement of the surrounding tissues, and prevention of recurrent microbial soilage. In general, this requires draining, resecting, excluding, or patching the involved viscus during laparotomy or laparoscopy. Intraoperative saline lavage and radical peritoneal debridement have not proved useful; postoperative peritoneal lavage and planned repeat laparotomy have been suggested but are not of established benefit.

Approximately 20% to 30% of patients who require an operation for treatment of peritonitis or intra-abdominal abscess will require a second operative procedure to resolve the infection and establish adequate source control.27 Surgeons' assessments of the adequacy of the source control achieved with the original operative procedure have been found to be strongly predictive of patients' subsequent need for reoperation and mortality.28

The choice of antimicrobial agent depends on the organisms involved in the peritonitis. Initial selection, however, is always made before culture results are available, and it must take into consideration the organisms that are predominant in the colon: B. fragilis, enteric gram-negative bacilli, streptococci, and enterococci. Animal studies and clinical experience have shown the importance of using antibiotics that are effective against both aerobic and anaerobic bacteria to treat patients with polymicrobial peritonitis, but clinical trials have failed to establish the superiority of any particular regimen. Newer antibiotics that provide broad-spectrum coverage of many aerobic and anaerobic species allow single antibiotic therapy in many cases. Useful single agents include ampicillin-sulbactam, ticarcillin-clavulanic acid, cefoxitin, cefotetan, piperacillin-tazobactam, ertapenem, meropenem, and imipenem-cilastatin. Effective multidrug regimens include (1) an aminoglycoside combined with either clindamycin or metronidazole, (2) aztreonam combined with clindamycin, and (3) a quinolone or a third-generation cephalosporin combined with either clindamycin or metronidazole29,30 [see 7:XIV Chemotherapy of Infection]. Although multiple drugs provide broader antimicrobial coverage, they do not appear to be more effective than single-drug regimens. In all cases, the final choice of antibiotics should be determined by the results of culture and sensitivity testing and the clinical course.

Most antibiotics attain concentrations in ascitic fluid that are at least half of the simultaneous serum levels and that exceed the minimum inhibitory concentration for the infecting organism. For this reason, systemic therapy alone is generally adequate in the management of bacterial peritonitis in patients with ascites; intraperitoneal instillation of antibiotics does not appear necessary. The necessary duration of antibiotic administration has never been systematically studied. For most patients, antimicrobial agents can be stopped as soon as clinical signs of infection begin to resolve, intestinal function resumes, and temperature and white blood cell count begin to return to normal. It is unusual for patients to require more than 7 days of treatment, and many patients can be managed with less.29,30



Infection continues to be a significant problem for peritoneal dialysis patients. Peritonitis develops in as many as 60% of patients undergoing continuous ambulatory peritoneal dialysis during the first year of treatment, and the infection recurs in 20% to 30% of these patients; elderly patients are the most vulnerable.31


Clinical features

The development of fever, abdominal pain or tenderness, and leukocytosis and the isolation of a bacterial or mycotic agent from the effluent fluid in a patient on peritoneal dialysis indicate peritonitis. Isolation of bacteria from the dialysate in a patient without these findings often signals contamination rather than infection. Turbidity of the dialysate from neutrophils occurs in 2% to 3% of dialyses. Although turbidity itself does not necessarily indicate peritonitis, it should be considered an indication of infection until the results of the culture are available. Absence of bacteria on a Gram stain of the dialysate sediment does not necessarily confirm the absence of infection, however, because of the extensive dilutions required. Therefore, one cannot rely on a negative Gram stain to discriminate between infection and sterile inflammation.31


The principal organisms in peritonitis that complicate peritoneal dialysis are coagulase-negative staphylococci, S. aureus, Pseudomonas aeruginosa, E. coli and other enteric organisms, and Candida species.31,32,33 Microorganisms may enter the peritoneal cavity exogenously (i.e., after colonization of the abdominal wound area or by contamination of the dialysate) or endogenously (i.e., by bacteremia or by transmural migration of bowel flora, perhaps enhanced by catheter-induced trauma). Most episodes are monomicrobial, but polymicrobial peritonitis can occur.34,35 Failure to respond to antibiotic therapy within 96 hours often signals infection with gram-negative bacilli (typically, P. aeruginosa); the prognosis in these patients is worse than in those who respond rapidly. Removal of the dialysis catheter may be necessary to control infection.


Peritonitis caused by Candida species occurs most often as a complication of peritoneal dialysis, gastrointestinal surgery, or perforation of an abdominal viscus. Candidal peritonitis that complicates peritoneal dialysis is treated with intravenous amphotericin B, intraperitoneal amphotericin B, or both at a final dialysate concentration of 2 to 4 µg/ml. Fluconazole may also be useful for treatment of peritonitis caused by C. albicans, as may caspofungin.31,36

Whereas the addition of peritoneal lavage, with or without antibiotics, does not appear to improve on the results achieved by intravenous antibiotics and conventional surgical therapy, intraperitoneal administration of antibiotics may be useful in patients who require peritoneal dialysis. For example, different antibiotics can be added directly to the dialysate in specific concentrations, such as ampicillin in a concentration of 50 mg/L or gentamicin in a concentration of 5 to 10 mg/L. Because bacteremia may occur, antibiotics should also be administered intravenously in these patients in a dosage appropriate to the patient's level of renal function. When peritonitis develops as a complication of peritoneal dialysis or peritoneovenous shunting, it is often necessary to remove or replace the catheter during administration of antibiotics to control the peritonitis.31,36 In patients with a history of peritonitis caused by S. aureus, prophylaxis with either topical mupirocin ointment in the nares or oral rifampin may reduce the incidence of subsequent episodes of staphylococcal peritonitis and peritoneal catheter loss.31,36,37

Intra-abdominal Abscesses

Intra-abdominal abscesses may present as complications of an abdominal operation, intra-abdominal conditions (e.g., diverticulitis, appendicitis, biliary tract disease, pancreatitis, perforated viscus, peritonitis), or penetrating abdominal trauma; as fever of obscure origin; or as dysfunction of neighboring organs (e.g., so-called lower lobe pneumonia related to a subphrenic abscess or small bowel obstruction). Bacteremic spread of infection from a distant focus to an intra-abdominal site is a less common cause of intra-abdominal abscesses.

Intra-abdominal abscesses are conveniently classified according to the anatomic location in which they occur: intraperitoneal, retroperitoneal, or visceral. Intraperitoneal abscesses are areas of localized peritonitis in which infection has progressed but has been walled off by omentum, peritoneum, and adjacent viscera. Retroperitoneal infections include pancreatitis-associated infections, perinephric abscesses, and paravertebral abscesses. Visceral abscesses develop within abdominal viscera—predominantly, the liver and, less often, the spleen—and other organs. In general, the location of the abscess does not affect the diagnosis or treatment beyond influencing the choice of percutaneous or surgical drainage.



Although the location of the abscess determines its particular features, many intra-abdominal infections share common elements. Fever, for example, is almost invariable; it often recurs in a spiking pattern and may be accompanied by rigors. Hypotension and even septic shock may develop. Abdominal pain is a major clue to the presence of an intra-abdominal abscess: when present, it can predominate; when it is absent, the diagnosis can be very difficult. Geriatric patients, in particular, may present atypically, without abdominal pain or fever. On laboratory studies, leukocytosis and elevation of liver enzyme and serum amylase levels are common. Bacteremia, which may be polymicrobial, occurs in up to one fourth of cases.

Intra-abdominal abscesses characteristically contain multiple bacterial species. Anaerobic bacteria can be isolated from 60% to 70% of such abscesses; the bacteria most commonly isolated include B. fragilis, peptostreptococci and peptococci, Clostridium species, and facultative species such as E. coli, Enterobacter, Klebsiella, and enterococci. The specific organisms isolated do not generally provide clues to the nature of the underlying process. However, the presence of Citrobacter species strongly suggests a biliary or upper gastrointestinal source;S. aureus, otherwise uncommon in intra-abdominal abscesses, suggests bacteremic seeding or vertebral osteomyelitis, which can lead to retroperitoneal or perinephric abscesses.

Plain radiography (kidneys, ureters, and bladder [KUB]; upright and lateral decubitus views) may afford important clues to the diagnosis of intra-abdominal abscesses. For example, air-fluid levels may indicate an intra-abdominal collection, free air may point to perforation of a viscus as the underlying problem, displaced loops of bowel may signify an abscess, and a so-called soap-bubble appearance or loss of the normal psoas shadow may suggest a retroperitoneal collection.

Ultrasonography and computed tomographic scanning, however, are much more sensitive and specific than plain radiography and are now the standard radiologic techniques for evaluating intra-abdominal abscesses.38 Both are excellent for diagnosis, and both can be used to guide percutaneous abscess drainage [see Treatment and Prevention, below].39 CT scanning is the more accurate study; its specificity and sensitivity rates can exceed 90%. Compared with ultrasonography, CT scanning has the additional advantages of allowing simultaneous administration of contrast, of not requiring skin contact (hence, surgical dressings do not interfere with the study), and of producing accurate results even in the presence of ileus and abdominal gas collections. Ultrasonography, however, is less expensive, is often more readily available, can sometimes be done with portable equipment at the bedside, and does not involve exposure to radiation. Ultrasonography is most accurate for detecting abscesses in the left or right upper quadrant of the abdomen and in the true pelvis; it is also sensitive and specific for identifying ascites. In patients with acute abdominal disease, however, ultrasonography is often limited by bowel gas, which obscures any deeper findings.

Magnetic resonance imaging plays a negligible role in the evaluation of intra-abdominal infections. Nuclear medicine studies are also less helpful than CT and ultrasonography. Although early results appeared promising, gallium-67 scanning and indium-111 scanning both have proved less helpful than CT and ultrasonographic techniques. Cholescintigraphy scans using technetium-99m-labeled hepatoiminodiacetic acid (HIDA, or lidofenin) are useful in evaluating the gallbladder and for demonstrating a bile leak after cholecystectomy or other biliary procedure [see 4:VI Gallstones and Biliary Tract Disease]. Arteriography and barium contrast studies are seldom used to diagnose intra-abdominal abscesses. If fistulous tracts are present, however, sinograms may occasionally be helpful.

Treatment and Prevention

The choice of antibiotics depends on the organisms isolated from cultures of blood or abscess material. Until this information is available, the choice of drugs should be guided by the same principles as those that apply to the treatment of peritonitis. Although the use of antibiotics is essential, especially because of the risk of bacteremia, such therapy alone will not eradicate intra-abdominal abscesses and is therefore secondary to prompt, effective abscess drainage.

Until the mid-1970s, surgical drainage was mandatory for the treatment of intra-abdominal abscesses. However, treatment changed dramatically within just a few years of the introduction of percutaneous abscess drainage under ultrasonographic or CT guidance. Ultrasonography may be used to guide drainage of large or superficial collections, but CT is preferable for smaller or deeper abscesses.40Many studies have demonstrated that percutaneous abscess drainage is safe and effective for a broad range of intra-abdominal collections; success rates range from 47% to 92%, with most studies reporting better than 80% success, similar to the success rate for surgical drainage.41 Failure of treatment is more common in immunosuppressed patients and in those with poorly defined phlegmons, multilocular abscesses, thick hematomas or organized infections, or abscesses with associated fistulous tracts.

Radiographic features alone cannot indicate which abscesses will respond to percutaneous drainage. Hence, it seems reasonable to institute percutaneous drainage in all patients who have a safe access route, provided that skilled personnel are available and that the patient does not otherwise require surgical intervention. Surgical drainage may then be used in patients with recurrences, failures, or complications. A surgeon should be involved in the decision regarding the method of drainage, because the surgeon will be called if the initial approach is not successful. Even in the case of abscesses that usually require surgical intervention, such as periappendiceal and diverticular abscesses and peripancreatic infections (see below), percutaneous drainage can provide temporary control of sepsis, allowing the operative procedure to be delayed until conditions are optimal and sometimes allowing a single definitive procedure instead of staged procedures.

A number of effective preventive strategies are available to reduce the likelihood of both wound infections (incisional surgical site infections) and intra-abdominal abscesses (organ/space surgical site infections) after abdominal operations.42 These include the appropriate use of prophylactic antibiotics,43 maintenance of normothermia in the operating room,44,45 provision of high levels of inspired oxygen,45,46adequate fluid resuscitation during the operation,47 and maintenance of euglycemia in the perioperative period.48,49,50


Intraperitoneal abscesses may form in either of two ways: (1) from diffuse peritonitis in which loculations of pus develop in anatomically dependent areas such as the pelvis, paracolic gutters, and subphrenic areas or (2) by spread of infection from a localized inflammatory process to contiguous peritoneum. About one third of intra-abdominal abscesses are intraperitoneal, and almost one half of intraperitoneal abscesses occur in the right lower quadrant.

Subphrenic Abscesses

About 60% of subphrenic abscesses develop after operations involving the duodenum and stomach, biliary tract, or appendix; 20% to 40% develop after rupture of a hollow viscus (such as perforated peptic ulcer or acute appendicitis), in which the infection is subsequently sealed off. A variable percentage of subphrenic abscesses develop after penetrating or blunt (closed) abdominal trauma, and less than 5% develop without predisposing circumstances. Diagnosis of subphrenic abscesses is sometimes delayed because of their location in the intrathoracic portion of the peritoneal cavity, which is not amenable to examination.

Clinical features

The manifestations of a subphrenic abscess range from a severe acute illness to an insidious chronic process characterized by intermittent fever, weight loss, anemia, and nonspecific symptoms. The chronic syndrome is most often observed in patients who have previously received antibiotics; in the past, such an abscess could smolder subclinically for prolonged periods before diagnosis. This is currently uncommon. In any patient with fever of undetermined origin who has had an abdominal operation—even if the operation was performed many months earlier—a chronic intra-abdominal abscess must be suspected and a CT scan should be done.

Spiking fever, abdominal pain and tenderness (most often at the lower costal margin), and weight loss are common manifestations. Features of an intrathoracic process, such as shoulder pain, chest pain, cough, dyspnea, rales, and pleural effusion, are more commonly observed than features of an intra-abdominal condition. Leukocytosis is common. Rarely, patients will have a prolonged, obscure febrile illness complicated by the sudden development of an empyema when the subphrenic abscess ruptures through the diaphragm. Although pleural fluid is present in about 80% of patients with a subphrenic abscess, it is usually a sympathetic transudate. A pleural effusion that develops after an abdominal operation is more commonly caused by inflammation below the diaphragm than inflammation above it.


CT scanning and ultrasonography are the best radiologic techniques for establishing the diagnosis. The plain x-ray findings in patients with subphrenic abscess include pleural effusion, limitation of diaphragmatic movement, elevation of a hemidiaphragm, and lower-lobe pneumonia or atelectasis.


Pyogenic infections of the retroperitoneum present like other intra-abdominal infections. Indeed, many retroperitoneal abscesses arise from disorders of the abdominal viscera; more than two thirds of patients with retroperitoneal abscesses also have underlying debilitating conditions, including malignancies, corticosteroid use, alcoholism, and diabetes. More than 80% of these infections are polymicrobial, involving aerobic and anaerobic enteric organisms.51 CT scanning is the key to diagnosis of retroperitoneal abscesses. The same is true for primary psoas abscesses, which are often caused by S. aureus,52 and for perinephric abscesses, which usually originate in the urinary tract.53 As with other abscesses, successful management requires prompt percutaneous or surgical drainage and the administration of appropriate antibiotics.53,54

Pancreatic Infections

Most pancreatic infections occur as a complication of pancreatitis, which can result from alcoholism (38%), gallstones (11%), surgical trauma (16%), or other factors (35%). Pancreatic infections have been divided into infected pancreatic and peripancreatic necrosis and pancreatic abscesses.55 Infected pancreatic necrosis tends to occur during the first 3 weeks after the onset of acute necrotizing pancreatitis and is poorly localized within the retroperitoneum. Source control is difficult, and morbidity and mortality can be high. Infected necrosis often requires open operative debridement, although there are promising reports of combined percutaneous drainage and laparoscopically assisted debridement. Pancreatic abscess refers to a more localized infectious process. It can occur either in or adjacent to the pancreas and tends to occur more than 3 weeks after the acute onset of disease. An infected pseudocyst is also a pancreatic abscess. Pancreatic abscesses can often be managed percutaneously but are still likely to require additional source control.56

Pancreatic infections are often polymicrobial, typically containing three or four species of bacteria. Most are enteric organisms, including E. coli, enterococci, Klebsiella species, and anaerobes such as Bacteroides, Peptococcus, Fusobacterium, and Clostridium species. Nonenteric organisms, including staphylococci, P. aeruginosa, and, less often, Candida species, may be involved. Bacteremia occurs in about 26% of cases. Recent reports demonstrate a shift in the usual microbial flora of infected pancreatitis, with an increase in gram-positive cocci and fungi (e.g., Candida species), probably secondary to the increasingly common use of prophylactic antibiotics for long periods in patients with necrotizing pancreatitis.57,58

Clinical features

The initial presentation of noninfected acute necrotizing pancreatitis involves fever, leukocytosis, and abdominal pain and tenderness. The clinical features do not allow the differentiation between infected and uninfected patients. Most infections occur after at least 1 week of disease.


The most accurate method for determining whether an area of pancreatic or peripancreatic necrosis is infected is to perform CT scan-guided fine-needle aspiration for Gram stain and culture. This step is indicated if a patient's clinical condition deteriorates after initial stabilization or improvement.59


Source control is mandatory. It can often be accomplished by open surgical debridement, sometimes aided by percutaneous drainage, laparoscopic techniques, or both. The outcome is improved when intervention occurs later in the course of the disease. Antibiotic therapy is the same as for other intra-abdominal infections.


Liver Abscesses

Epidemiology and etiology

Pyogenic liver abscesses occur in several settings, including biliary tract infection, direct extension from a contiguous site of infection, portal bacteremia from intra-abdominal septic foci, and nonpenetrating trauma.60,61,62 Liver abscesses may occur as a result of systemic bacteremia or as complications of abdominal surgery or penetrating abdominal trauma. They may also occur as complications of hepatocellular carcinoma,63 chronic granulomatous disease,64,65 or percutaneous transhepatic biliary drainage procedures in patients with cancer and obstructive jaundice. Pyogenic abscesses may be single or multiple.

Like other intra-abdominal abscesses, pyogenic liver abscesses principally involve enteric bacteria; two thirds of these abscesses have polymicrobial origins, and at least one third involve anaerobes. S. aureus may be the causative organism in patients with bacteremia and in children. Klebsiella species are often responsible for gas-forming liver abscesses, which typically occur in patients with diabetes.66 Blood cultures are positive in about half of patients with a pyogenic liver abscess, and metastatic infections may occur.

Clinical features

Fever is the most common symptom and is present in nearly 90% of patients. Chills and weight loss occur in about half of cases. Because abdominal pain, abdominal tenderness, or hepatomegaly is present in only half of cases, many of these patients present with fever of undetermined origin. Leukocytosis is present in most cases. Jaundice is infrequent, but the serum alkaline phosphatase level is elevated in almost all patients. Rupture of a liver abscess, although uncommon, is often accompanied by diffuse abdominal pain and septic shock.67


CT scanning is the most accurate diagnostic technique [see Figures 1a and 1b], yielding positive results in up to 95% of confirmed cases; ultrasonography is also helpful, yielding positive results in up to 80% of confirmed cases. The initial clue to the diagnosis may come from plain x-rays, which may show an elevated right hemidiaphragm, a right pleural effusion, or an air-fluid level.


Figure 1a. CT Scan: Multilobular Liver Abscess

CT scan shows multilobular liver abscess (arrow).


Figure 1b. CT Scan: Resolution of Abscess Cavity

Four days after percutaneous abscess drainage, CT scan shows resolution of the abscess cavity.

The major differential diagnosis is amebic liver abscess [see 7:XXXIV Protozoan Infections]. Amebic abscesses are more likely to be solitary and confined to the right lobe of the liver; a history of travel or diarrhea may suggest the diagnosis. Stool ova and parasite examination revealing Entamoeba histolytica is highly suggestive, but results are often negative in patients with hepatic amebiasis. However, most patients with amebic liver abscesses have positive amebic serologies. It is important to note that E. histolytica has been reclassified into two morphologically similar but genetically distinct species: E. histolytica, the pathogenic protozoan that causes amebic dysentery and hepatic abscess, and E. dispar, a nonpathogenic commensal protozoan of humans. Specific enzyme-linked immunosorbent assay-based serologic testing to distinguish E. dispar colonization from E. histolytica infection is recommended before treating amebiasis.68,69 Less often, hepatic cysts or neoplasms may be confused with liver abscesses.


Whereas surgery was formerly the mainstay of therapy, percutaneous drainage should now be the initial drainage procedure in most patients with pyogenic liver abscesses. Antibiotics with broad coverage of enteric organisms and staphylococci should be administered intravenously until specific pathogens have been isolated from the abscess or the bloodstream.70 Mortality depends largely on the underlying disease and is highest in patients with cancer.71 Surgical therapy is required for ruptured abscesses, but the mortality is high, approaching 44%.67

Splenic Abscesses

Splenic abscesses are uncommon.72 Unlike other intra-abdominal abscesses, they are often bacteremic in origin, especially in patients with endocarditis. In other patients, hemoglobinopathy, vasculitis with splenic infarction, trauma, and immunosuppression may be predisposing factors. Fever and chills and left upper quadrant pain are common. If the upper pole of the spleen is affected, diaphragmatic and pleural and pulmonary symptoms may predominate, but peritoneal symptoms are more common if the lower pole is the site of infection. Responsible organisms include S. aureus, streptococci, Salmonella species, and enteric bacteria; fungi are important causes in immunocompromised patients. CT scans and ultrasonography are the most useful radiographic studies. Appropriate antimicrobial therapy is essential. Although splenectomy has often been required for effective management in the past, evidence now indicates that percutaneous drainage or even antibiotics alone may suffice in selected cases.73 More experience is needed before the optimal management of these uncommon infections can be determined.74


  1. Johnson CC, Baldessarre J, Levison ME: Peritonitis: update on pathophysiology, clinical manifestations, and management. Clin Infect Dis 24:1035, 1997
  2. Hall JC, Heel KA, Papadimitriou JM, et al: The pathology of peritonitis. Gastroenterology 114:185, 1998
  3. Such J, Runyon BA: Spontaneous bacterial peritonitis. Clin Infect Dis 27:669, 1998
  4. Dhiman RK, Makharia GK, Jain S, et al: Ascites and spontaneous bacterial peritonitis in fulminant hepatic failure. Am J Gastroenterol 95:233, 2000
  5. Thuluvath PJ, Morss S, Thompson R: Spontaneous bacterial peritonitis: in-hospital mortality, predictors of survival, and health care costs from 1988 to 1998. Am J Gastroenterol 96:1232, 2001
  6. Garcia-Tsao G: Currrent management of the complications of cirrhosis and portal hypertension: variceal hemorrhage, ascites, and spontaneous bacterial peritonitis. Gastroenterology 120:726, 2001
  7. Rimola A, Garcia-Tsao G, Navasa M, et al: Diagnosis, treatment, and prophylaxis of spontaneous bacterial peritonitis: a consensus document. International Ascites Club. J Hepatol 32:142, 2000
  8. Guarner C, Sola R, Soriano G, et al: Risk of a first community-acquired spontaneous bacterial peritonitis in cirrhotics with low ascitic fluid protein levels. Gastroenterology 117:414, 1999
  9. Yoneyama K, Miyagishi K, Kiuschi Y, et al: Risk factors for infections in cirrhotic patients with and without hepatocellular carcinoma. J Gastroenterol 37:1028, 2002
  10. Hemsley C, Eykyn SJ: Pneumococcal peritonitis in previously healthy adults: case report and review. Clin Infect Dis 27:376, 1998
  11. Hillebrand DJ, Runyon BA: Spontaneous bacterial peritonitis: keys to management. Hosp Pract (Off Ed) 35:87, 2000
  12. Llovet JM, Moitinho E, Sala M, et al: Prevalence and prognostic value of hepatocellular carcinoma in cirrhotic patients presenting with spontaneous bacterial peritonitis. J Hepatol 33:423, 2000
  13. Hampel H, Bynum GD, Zamora E, et al: Risk factors for the development of renal dysfunction in hospitalized patients with cirrhosis. Am J Gastroenterol 96:2206, 2001
  14. Franca A, Giordano HM, Seva-Pereira T, et al: Five days of ceftriaxone to treat spontaneous bacterial peritonitis in cirrhotic patients. J Gastroenterol 37:119, 2002
  15. Runyon BA: Albumin infusion for spontaneous bacterial peritonitis. Lancet 354:1838, 1999
  16. Sort P, Navasa M, Arroyo V, et al: Effect of intravenous albumin on renal impairment and mortality in patients with cirrhosis and spontaneous bacterial peritonitis. N Engl J Med 341:403, 1999
  17. Grange JD, Roulot D, Pelletier G, et al: Norfloxacin primary prophylaxis of bacterial infections in cirrhotic patients with ascites: a double-blind randomized trial. J Hepatol 29:430, 1998
  18. Gines P, Navasa M: Antibiotic prophylaxis for spontaneous bacterial peritonitis: how and whom? J Hepatol 29:490, 1998
  19. Bernard B, Grange JD, Khac EN, et al: Antibiotic prophylaxis for the prevention of bacterial infections in cirrhotic patients with gastrointestinal bleeding: a meta-analysis. Hepatology 29:1655, 1999
  20. Soares-Weiser K, Brezis M, Tur-Kaspa R, et al: Antibiotic prophylaxis for cirrhotic patients with gastrointestinal bleeding. Cochrane Database Syst Rev 2:CD002907, 2002
  21. Seiler CA, Brugger L, Forssmann U, et al: Conservative surgical treatment of diffuse peritonitis. Surgery 127:178, 2000
  22. Wilson SE, Hopkins JA: Clinical correlates of anaerobic bacteriology in peritonitis. Clin Infect Dis 20(suppl 2):S251, 1995
  23. Barie PS, Vogel SB, Dellinger EP, et al: A randomized, double-blind clinical trial comparing cefepime plus metronidazole with imipenem-cilastatin in the treatment of complicated intra-abdominal infections. Cefepime Intra-abdominal Infection Study Group. Arch Surg 132:1294, 1997
  24. Dellinger EP, Wertz MJ, Meakins JL, et al: Surgical infection stratification system for intra-abdominal infection: multicenter trial. Arch Surg 120:21, 1985
  25. Lennard ES, Minshew BH, Dellinger EP, et al: Stratified outcome comparison of clindamycin-gentamicin vs chloramphenicol-gentamicin for treatment of intra-abdominal sepsis. Arch Surg 120:889, 1985
  26. Nathens AB, Rotstein OD, Marshall JC: Tertiary peritonitis: clinical features of a complex nosocomial infection. World J Surg 22:158, 1998
  27. Dellinger EP: Surgery for intra-abdominal sepsis. Reoperative Gastrointestinal Surgery. White TT, Harrison RC, Debas HT, et al, Eds. Appleton Century Crofts, New York, 1989, p 63
  28. Koperna T, Schulz F: Relaparotomy in peritonitis: prognosis and treatment of patients with persisting intraabdominal infection. World J Surg 24:32, 2000
  29. Mazuski JE, Sawyer RG, Nathens AB, et al: The Surgical Infection Society guidelines on antimicrobial therapy for intra-abdominal infections: an executive summary. Surg Infect (Larchmt) 3:161, 2002
  30. Mazuski JE, Sawyer RG, Nathens AB, et al: The Surgical Infection Society guidelines on antimicrobial therapy for intra-abdominal infections: evidence for the recommendations. Surg Infect (Larchmt) 3:175, 2002
  31. Vas S, Oreopoulos DG: Infections in patients undergoing peritoneal dialysis. Infect Dis Clin North Am 15:743, 2001
  32. Zelenitsky S, Barns L, Findlay I, et al: Analysis of microbiological trends in peritoneal dialysis-related peritonitis from 1991 to 1998. Am J Kidney Dis 36:1009, 2000
  33. Bibashi E, Memmos D, Kokolina E, et al: Fungal peritonitis complicating peritoneal dialysis during an 11-year period: report of 46 cases. Clin Infect Dis 36:927, 2003
  34. Kim GC, Korbet SM: Polymicrobial peritonitis in continuous ambulatory peritoneal dialysis patients. Am J Kidney Dis 36:1000, 2000
  35. Szeto CC, Chow KM, Wong TY, et al: Conservative management of polymicrobial peritonitis complicating peritoneal dialysis: a series of 140 consecutive cases. Am J Med 113:728, 2002
  36. Keane WF, Bailie GR, Boeschoten E, et al: Adult peritoneal dialysis-related peritonitis treatment recommendations: 2000 update. Perit Dial Int 20:396, 2000
  37. Ritzau J, Hoffman RM, Tzamaloukas AH: Effect of preventing Staphylococcus aureuscarriage on rates of peritoneal catheter-related staphylococcal infections: literature synthesis. Perit Dial Int 21:471, 2001
  38. Gupta H, Dupuy DE: Advances in imaging of the acute abdomen. Surg Clin North Am 77:1245, 1997
  39. Sheafor DH, Paulson EK, Simmons CM, et al: Abdominal percutaneous interventional procedures: comparison of CT and US guidance. Radiology 207:705, 1998
  40. Montgomery RS, Wilson SE: Intraabdominal abscesses: image-guided diagnosis and therapy. Clin Infect Dis 23:28, 1996
  41. Levison MA: Percutaneous versus open operative drainage of intra-abdominal abscesses. Infect Dis Clin North Am 6:525, 1992
  42. Horan TC, Gaynes RP, Martone WJ, et al: CDC definitions of nosocomial surgical site infections, 1992: a modification of CDC definitions of surgical wound infections. Am J Infect Control 20:271, 1992
  43. Dellinger EP, Gross PA, Barrett TL, et al: Quality standard for antimicrobial prophylaxis in surgical procedures. Infectious Diseases Society of America. Clin Infect Dis 18:422, 1994
  44. Kurz A, Sessler DI, Lenhardt R: Perioperative normothermia to reduce the incidence of surgical-wound infection and shorten hospitalization: study of wound infection and temperature group. N Engl J Med 334:1209, 1996
  45. Sessler DI, Akca O: Nonpharmacological prevention of surgical wound infections. Clin Infect Dis 35:1397, 2002
  46. Greif R, Akca O, Horn EP, et al: Supplemental perioperative oxygen to reduce the incidence of surgical-wound infection. Outcomes Research Group. N Engl J Med 342:161, 2000
  47. Arkilic CF, Taguchi A, Sharma N, et al: Supplemental perioperative fluid administration increases tissue oxygen pressure. Surgery 133:49, 2003
  48. Pomposelli JJ, Baxter JK 3rd, Babineau TJ, et al: Early postoperative glucose control predicts nosocomial infection rate in diabetic patients. J Parenter Enteral Nutr 22:77, 1998
  49. van den Berghe G, Wouters P, Weekers F, et al: Intensive insulin therapy in the surgical intensive care unit. N Engl J Med 345:1359, 2001
  50. Latham R, Lancaster AD, Covington JF, et al: The association of diabetes and glucose control with surgical-site infections among cardiothoracic surgery patients. Infect Control Hosp Epidemiol 22:607, 2001
  51. Brook I, Frazier EH: Aerobic and anaerobic microbiology of retroperitoneal abscesses. Clin Infect Dis 26:938, 1998
  52. Santaella RO, Fishman EK, Lipsett PA: Primary vs secondary iliopsoas abscess. Arch Surg 130:1309, 1995
  53. Dembry LM, Andriole VT: Renal and perirenal abscesses. Infect Dis Clin North Am 11:663, 1997
  54. Meng MV, Mario LA, McAninch JW: Current treatment and outcomes of perinephric abscesses. J Urol 168:1337, 2002
  55. Bradley EL: A clinically based classification system for acute pancreatitis. Summary of the International Symposium on Acute Pancreatitis, Atlanta, Ga, September 11 through 13, 1992. Arch Surg 128:586, 1993
  56. Dellinger P: Pancreatic infection.Source Control: A Guide to the Management of Surgical Infections. Schein M, Marshall JC, Eds. Springer, Berlin, Germany, 2002, p 154
  57. Fernandez-del Castillo C, Rattner DW, Makary MA, et al: Debridement and closed packing for the treatment of necrotizing pancreatitis. Ann Surg 228:676, 1998
  58. Howard TJ, Temple MB: Prophylactic antibiotics alter the bacteriology of infected necrosis in severe acute pancreatitis. J Am Coll Surg 195:759, 2002
  59. Buchler MW, Gloor B, Muller CA, et al: Acute necrotizing pancreatitis: treatment strategy according to the status of infection. Ann Surg 232:619, 2000
  60. Johannsen EC, Sifri CD, Madoff LC: Pyogenic liver abscesses. Infect Dis Clin North Am 14:547, 2000
  61. Mohsen AH, Green ST, Read RC, et al: Liver abscess in adults: ten years experience in a UK centre. QJM 95:797, 2002
  62. Ralls PW: Inflammatory disease of the liver. Clin Liver Dis 6:203, 2002
  63. Rintoul R, O'Riordan MG, Laurenson IF, et al: Changing management of pyogenic liver abscess. Br J Surg 83:1215, 1996
  64. Winkelstein JA, Marino MC, Johnston RB Jr, et al: Chronic granulomatous disease: report on a national registry of 368 patients. Medicine (Baltimore) 79:155, 2000
  65. Lublin M, Bartlett DL, Danforth DN, et al: Hepatic abscess in patients with chronic granulomatous disease. Ann Surg 235:383, 2002
  66. Chou F-F, Sheen-Chen SM, Chen YS: The comparison of clinical course and results of treatment between gas-forming and non-gas-forming pyogenic liver abscess. Arch Surg 130:401, 1995
  67. Chou F-F, Sheen-Chen SM, Lee TY: Rupture of pyogenic liver abscess. Am J Gastroenterol 90:767, 1995
  68. Hughes MA, Peri WA Jr: Amebic liver abscess. Infect Dis Clin North Am 14:565, 2000
  69. Stanley SL Jr: Amoebiasis. Lancet 361:1025, 2003
  70. Alvarez Perez JA, Gonzalez JJ, Baldonedo RF, et al: Clinical course, treatment, and multivariate analysis of risk factors for pyogenic liver abscess. Am J Surg 181:177, 2001
  71. Yeh TS, Jan YY, Jen LB, et al: Hepatocellular carcinoma presenting as pyogenic liver abscess: characteristics, diagnosis, and management. Clin Infect Dis 26:1224, 1998
  72. Ooi LL, Leong SS: Splenic abscesses from 1987 to 1995. Am J Surg 174:87, 1997
  73. Thanos L, Dailiana T, Papaioannou G, et al: Percutaneous CT-guided drainage of splenic abscess. AJR Am J Roentgenol 179:629, 2002
  74. Ng KK, Lee TY, Wan CF, et al: Splenic abscess: diagnosis and management. Hepatogastroenterology 49:567, 2002

Editors: Dale, David C.; Federman, Daniel D.