INTRODUCTION — Infections within the abdominal cavity typically arise because of inflammation or disruption of the gastrointestinal tract. Less commonly, they can arise from the gynecologic or urinary tract. Abdominal infections are usually polymicrobial and result in an intra-abdominal abscess or secondary peritonitis, which may be generalized or localized (phlegmon).
The approach to antimicrobial selection and administration for intra-abdominal infections in adults is discussed here. The general and surgical management of these infections are discussed in detail elsewhere. (See "Management of acute appendicitis in adults" and "Acute colonic diverticulitis: Medical management" and "Acute colonic diverticulitis: Surgical management" and "Acute cholangitis: Clinical manifestations, diagnosis, and management" and "Treatment of acute calculous cholecystitis" and "Acalculous cholecystitis: Clinical manifestations, diagnosis, and management" and "Overview of gastrointestinal tract perforation".)
The approach to management of abscesses within specific intra-abdominal organs (such as the liver or kidney) are also discussed in detail separately. (See "Pyogenic liver abscess" and "Invasive liver abscess syndrome caused by Klebsiella pneumoniae" and "Renal and perinephric abscess" and "Management and complications of tubo-ovarian abscess" and "Posthysterectomy pelvic abscess".)
Spontaneous peritonitis and peritonitis associated with peritoneal dialysis are also discussed elsewhere. (See "Spontaneous bacterial peritonitis in adults: Treatment and prophylaxis" and "Microbiology and therapy of peritonitis in peritoneal dialysis".)
MICROBIOLOGY — Intra-abdominal infections usually arise after a breach in the intrinsic mucosal defense barrier that allows normal bowel flora to inoculate the abdominal cavity. The precise microbiological spectrum depends on the precise gastrointestinal source (ie, small versus large bowel).
Colonic flora is especially common in intra-abdominal infections, reflecting the frequency of associated diseases at this anatomic site, including appendicitis, diverticulitis, carcinoma of the colon, inflammatory bowel disease, and previous colon surgery. Thus, the predominant bacteria involved in such infections are coliforms (mainly Escherichia coli, Klebsiella spp, Proteus spp, and Enterobacter spp) streptococci, enterococci, and anaerobic bacteria (picture 1). However, while colonic flora consists of approximately 400 species, an average of only four to six species are generally recovered from these intra-abdominal infections. The dominant isolates in most series are Bacteroides fragilis and E. coli (table 1) [1-6]. The probable factors contributing to this phenomenon include the limited ability of clinical laboratories to isolate all the different organisms as well as the ability of specific organisms to cause infection and survive based upon their virulence factors and capacity to adapt to new environmental conditions. For example, the capacity of B. fragilis to tolerate small amounts of oxygen contributes to its emergence as a highly invasive anaerobic pathogen in abdominal infections [7]. Experimental animal studies of intra-abdominal sepsis suggest that both anaerobes and coliforms contribute to the pathogenesis although they play different roles, with coliforms contributing to early sepsis and anaerobes implicated in the late sequelae with abscess formation [8]. (See "Pathophysiology, clinical clues, and recovery of organisms in anaerobic infections".)
Perforation of the proximal bowel, as with perforated peptic ulcer, results in an infection that is microbiologically distinct, reflecting the flora of the upper gastrointestinal tract. The predominant microbial species in such cases often include aerobic and anaerobic gram-positive bacteria or Candida spp. (See "Overview of complications of peptic ulcer disease", section on 'Perforation'.)
Prior antimicrobial therapy and health care exposures are associated with microbiologic changes in the bowel flora, and intra-abdominal infections in such settings are thus more likely to involve nosocomial pathogens, such as Pseudomonas aeruginosa and other drug-resistant organisms. Enterococci are most likely to be clinically relevant in health care-associated infections, particularly postoperative infections, in contrast to community-acquired infections, in which they are frequently isolated but are often not important pathogens [9,10]. Candida spp are also more common, in both small and large bowel processes, among patients with hospital-acquired infection, prior antibiotic exposure, immunocompromising conditions, or with recurrent infection [11]. (See 'Considerations for specific pathogens' below.)
SOURCE CONTROL AND DRAINAGE — Surgical intervention and/or percutaneous drainage are usually critical to the management of intra-abdominal infections other than spontaneous peritonitis. Surgical intervention may be required to close an anatomic breach or debride infected necrotic tissue, and drainage is usually necessary for clearance of an abscess. When feasible, percutaneous abscess drainage is preferred [11]. Most clinical treatment failures are due to failure to achieve such source control.
Surgical or percutaneous intervention also affords the opportunity for collection of primary specimens for microbiologic analysis (Gram stain, aerobic and anaerobic cultures, and if appropriate, fungal and mycobacterial studies). This is particularly important for those patients with intra-abdominal abscesses or otherwise complicated infections, with prior antibiotic exposure, or with a high risk of infection with resistant organisms. Gram stain of the specimen can provide early guidance for antibiotic selection and may be the only source of information if cultures do not grow. Inoculating the specimen directly into blood culture bottles can increase the microbiologic yield, but this approach has several drawbacks [11-13]. It forfeits the ability to obtain Gram stain results, unless a separate specimen is collected for Gram stain, and in polymicrobial infections, competitive growth in blood culture bottles can hinder identification of all important pathogens, so cultures with routine media are also important in such cases.
The surgical management of intra-abdominal processes is discussed in detail elsewhere. (See "Management of acute appendicitis in adults" and "Acute cholangitis: Clinical manifestations, diagnosis, and management", section on 'Biliary drainage' and "Treatment of acute calculous cholecystitis" and "Acalculous cholecystitis: Clinical manifestations, diagnosis, and management", section on 'Management' and "Overview of gastrointestinal tract perforation", section on 'Indications for abdominal exploration' and "Acute colonic diverticulitis: Surgical management".)
EMPIRIC ANTIMICROBIAL THERAPY
Timing — Patients who are critically ill should receive empiric antimicrobial therapy as soon as possible, ideally once blood and urine samples have been obtained for culture. In patients who are not critically ill, delaying antibiotic therapy until samples from the site of abdominal infection have been obtained for culture can be helpful to optimize the microbiologic yield that guides subsequent antibiotic selection.
Regimens — In general, empiric regimens for intra-abdominal infections include antimicrobial activity against enteric streptococci, coliforms, and anaerobes (table 2 and table 3 and table 4). Studies evaluating the relative efficacy of different antibiotic regimens with these spectra of activity have generally demonstrated equivalent efficacy (see 'General principles of regimen selection' below). The precise antimicrobial regimen and indications for broader antimicrobial coverage depend upon several factors (table 5):
●Whether the infection is community-acquired versus health care-associated.
●Whether there are individual risk factors for infection with resistant bacteria (such as recent travel to areas of the world that have high rates of antibiotic-resistant organisms or known colonization with such organisms).
●Whether the patient is considered to be at high risk for adverse outcomes. High-risk features that are associated with poor outcomes after intra-abdominal infection are advanced age (>70 years), delay in initial intervention >24 hours, inability to achieve adequate debridement or control of infection with drainage, other comorbidity (eg, renal or liver disease, malignancy), immunocompromising condition (eg, poorly controlled diabetes mellitus, chronic high-dose corticosteroid use, use of other immunosuppressive agents, neutropenia, advanced HIV infection, B or T lymphocyte defects), organ dysfunction, severe peritoneal involvement or diffuse peritonitis, low albumin level, and poor nutritional status [11,14-16].
Patients with community-acquired infections of mild to moderate severity who have none of these risk factors may not warrant very broad coverage, as the likelihood of resistant bacteria is low and the consequences of not covering them empirically are less. In contrast, broad coverage is appropriate in patients who are at risk for infection with resistant bacteria or who are at risk for adverse outcomes and mortality should empiric antibiotic therapy not be adequate. Thus, regimen selection is somewhat different for these different populations. (See 'Low-risk community-acquired infections' below and 'High-risk community-acquired infections' below and 'Health care-associated infections' below.)
Other factors that influence the choice of regimen include the location or type of infection (ie, gram-negative anaerobes are generally not critical pathogens in infections arising from the upper gastrointestinal tract), whether there is a plan for surgical intervention, the local rates of antibiotic-resistant Enterobacteriaceae, and expected patient tolerance. Rates of antibiotic resistance in Enterobacteriaceae are high in certain parts of the world, including east Asia, Africa, and the Middle East, and are especially high in southeast Asia [17]. Travelers from these areas are at risk for colonization with resistant bacteria; this risk generally lasts a few weeks but can be prolonged in those with diarrhea or antibiotic exposure during travel [17-20].
Overarching these considerations are goals for antibiotic stewardship, which generally favor narrower rather than broader coverage when possible.
The suggested regimens in the following sections are intended to provide general guidance and may need to be altered to cover emerging resistance patterns that are specific to the hospital or region; this is more likely to be necessary for the aerobic component but could also apply to the anaerobic component.
These recommendations are also generally in keeping with the joint Surgical Infection Society (SIS) and the Infectious Diseases Society of America (IDSA) guidelines on the management of complicated intra-abdominal infections, which were published in 2010. The SIS published updated guidelines in 2017, while the IDSA guidelines are in the process of revision [11,21]. There are some exceptions based on updated in vitro susceptibility data. As examples, although clindamycin and cefotetan were previously considered acceptable options for intra-abdominal infections involving anaerobes, these drugs are no longer recommended due to escalating rates of resistance in the B. fragilis group. As detailed in those guidelines, ampicillin-sulbactam is also not recommended due to high rates of resistance among community-acquired E. coli.
Low-risk community-acquired infections — For patients with mild to moderate community-acquired intra-abdominal infections (eg, perforated appendix or appendiceal abscess) who have no risk factors for antibiotic resistance or treatment failure (table 5), coverage of streptococci, nonresistant Enterobacteriaceae, and (in most cases) anaerobes is generally sufficient (table 2). The following initial empiric regimens are appropriate:
●Single-agent regimens – Piperacillin-tazobactam.
●Combination regimens – Cefazolin, cefuroxime, ceftriaxone, cefotaxime, ciprofloxacin, or levofloxacin, each in combination with metronidazole (although for most uncomplicated biliary infections of mild to moderate severity, the addition of metronidazole is not necessary).
When piperacillin-tazobactam or one of the above combination regimens cannot be used, ertapenem is a reasonable alternative but we try to preserve this agent for more resistant infections whenever possible.
An oral regimen (for example, a fluoroquinolone plus metronidazole or monotherapy with amoxicillin-clavulanic acid) is a reasonable choice for empiric therapy for patients with mild-moderate community-acquired infection who have no risk factors for infection with antibiotic-resistant organisms and when the prevalence of E. coli susceptibility to the chosen regimen exceeds 90 percent in the community and hospital. Oral regimens can still be used if the prevalence of E. coli susceptibility to oral options is less than 90 percent, but clinicians and patients should be aware of the greater risk of regimen failure in this case.
For these community-acquired infections, an empiric antimicrobial regimen does not have to include specific activity against enterococci or Pseudomonas. In several trials, clinical outcomes for community-acquired intra-abdominal infections have been similar with empiric regimens that have enterococcal and/or pseudomonal activity and those that do not [9,22-26].
The SIS/IDSA 2010 guidelines also list cefoxitin, moxifloxacin, and tigecycline as options, but we generally avoid using these agents in this setting [11]. This is because of substantial rates of in vitro resistance to cefoxitin and fluoroquinolones among Bacteroides spp and coliforms [27,28] and concern for increased mortality associated with tigecycline compared with other antibiotics for various infections, including intra-abdominal infections [29,30]. (See "Anaerobic bacterial infections", section on 'Antimicrobial resistance' and "Treatment of hospital-acquired and ventilator-associated pneumonia in adults", section on 'Other agents'.)
The SIS 2017 guidelines also list cefoperazone-sulbactam as an alternative, but we generally avoid using this agent in the interest of antimicrobial stewardship, as it provides coverage for resistant Pseudomonas spp that is typically not needed for this patient population [21]. These guidelines also suggest against cefazolin for empiric therapy of intra-abdominal infections because of the lack of trial data informing its use in such infections; we continue to use cefazolin as an option for low-risk community-acquired infections as long as the risk of resistance is not high (eg, local prevalence of cefazolin resistance in Enterobacteriaceae <10 percent, no recent antibiotic use).
High-risk community-acquired infections — For community-acquired intra-abdominal infections that are severe or in patients at high risk for adverse outcomes or resistance (table 5), broader coverage is warranted in an attempt to minimize the risk of inadequate empiric treatment. We generally include an agent with gram-negative activity broad enough to cover P. aeruginosa and Enterobacteriaceae that are resistant to nonpseudomonal cephalosporins in addition to coverage against enteric streptococci and (in most cases) anaerobes (table 3). Empiric antifungal therapy is usually not warranted, but it is reasonable for critically ill patients with an upper gastrointestinal source [21]. For community-acquired infections that clearly have an abdominal source, coverage for MRSA is generally not warranted, even in those individuals known to be MRSA-colonized.
The following initial empiric regimens are appropriate in areas where the local rates of resistance to these antibiotics are <10 percent:
●Single-agent regimens – Piperacillin-tazobactam.
●Combination regimens – Cefepime or ceftazidime, each administered with metronidazole.
If the patient cannot tolerate beta-lactams or is at risk for infection with an extended-spectrum beta-lactamase (ESBL)-producing organism (eg, known colonization or prior infection with an ESBL-producing organism), a carbapenem (imipenem or meropenem) should be chosen. The SIS 2017 guidelines also recommend adding vancomycin or ampicillin for regimens other than imipenem or piperacillin-tazobactam to provide enterococcal coverage, but we do not routinely employ empiric coverage of Enterococcus spp for community-acquired infections.
For patients who cannot use beta-lactams or carbapenems (eg, because of severe reactions), vancomycin plus aztreonam plus metronidazole is an alternative regimen.
For critically ill patients who warrant empiric antifungal therapy, an echinocandin (eg, anidulafungin, caspofungin, micafungin) is appropriate. (See "Management of candidemia and invasive candidiasis in adults".)
When beta-lactams or carbapenems are chosen for patients who are critically ill or are at high risk of infection with drug-resistant pathogens, we favor a prolonged infusion dosing strategy, which is also endorsed by the World Society of Emergency Surgery (WSES) [31]. (See "Prolonged infusions of beta-lactam antibiotics".)
Health care-associated infections — For patients with health care-associated infections, the likelihood of drug resistance is high. Thus, to achieve empiric coverage of likely pathogens, in addition to coverage against streptococci and anaerobes, regimens should at least include agents with expanded spectra of activity against gram-negative bacilli (including P. aeruginosa and Enterobacteriaceae that are resistant to nonpseudomonal third-generation cephalosporins and fluoroquinolones). We also usually use an empiric regimen that has anti-enterococcal activity for patients with health care–associated intra-abdominal infection, particularly those with postoperative infection, those who have previously received cephalosporins or other antimicrobial agents selecting for Enterococcus species, immunocompromised patients, and those with valvular heart disease or prosthetic intravascular materials (table 4).
Single-drug regimens that have expanded activity against gram-negative aerobic and anaerobic bacilli include meropenem, imipenem, and piperacillin-tazobactam. Combination regimens include ceftazidime or cefepime plus metronidazole. The combination of vancomycin, aztreonam, and metronidazole is an alternative for those who cannot use beta-lactams or carbapenems (eg, because of severe reactions). Cephalosporin-based regimens lack anti-enterococcal activity, so ampicillin or vancomycin can be added to these regimens for enterococcal coverage until culture results are available. When beta-lactams or carbapenems are chosen for patients who are critically ill or are at high risk of infection with drug-resistant pathogens, we favor a prolonged infusion dosing strategy, if feasible. (See "Prolonged infusions of beta-lactam antibiotics".)
Additions or modifications to the regimen may be indicated if other risk factors are present:
●For patients with health care-associated intra-abdominal infection who are known to be colonized with MRSA or who are at risk of having an infection due to this organism because of prior treatment failure and significant antibiotic exposure, empiric antimicrobial coverage directed against MRSA, typically with vancomycin, should be provided. (See "Methicillin-resistant Staphylococcus aureus (MRSA) in adults: Treatment of bacteremia".)
●For patients at risk for infection with an extended-spectrum beta-lactamase (ESBL) -producing organism (eg, known colonization or prior infection with an ESBL-producing organism), a carbapenem (imipenem or meropenem) should be chosen. (See "Extended-spectrum beta-lactamases", section on 'Preferred agents'.)
●For patients who are known to be colonized with highly resistant gram-negative bacteria, the addition of an aminoglycoside, polymyxin, or novel beta-lactam combination (ceftolozane-tazobactam or ceftazidime-avibactam) to an empiric regimen may be warranted. These agents should be combined with an agent that has activity against anaerobes (eg, metronidazole). (See "Principles of antimicrobial therapy of Pseudomonas aeruginosa infections", section on 'Management of multidrug-resistant organisms' and "Acinetobacter infection: Treatment and prevention", section on 'Second-line antibiotics' and "Overview of carbapenemase-producing gram-negative bacilli", section on 'Treatment'.)
●Empiric vancomycin-resistant Enterococcus (VRE) coverage is not generally recommended, except for patients who are at very high risk for infection due to VRE. These include liver transplant recipients with an intra-abdominal infection of hepatobiliary source and patients known to be colonized with VRE. In such cases, including a VRE-active agent (such as linezolid or daptomycin) in the empiric regimen is reasonable. [11]. For those known to be colonized with ampicillin-sensitive VRE, ampicillin, piperacillin-tazobactam, or imipenem can be used for coverage. (See "Treatment of enterococcal infections".)
●Empiric antifungal coverage is appropriate for patients at risk for infection with Candida spp, including those with upper gastrointestinal perforations, recurrent bowel perforations, surgically treated pancreatitis, heavy colonization with Candida spp, and/or yeast identified on Gram stain of samples from infected peritoneal fluid or tissue [21]. Fluconazole can be used for patients who are not severely ill and have no history of infection with a fluconazole-resistant isolate; otherwise, an echinocandin can be used. (See "Management of candidemia and invasive candidiasis in adults".)
TARGETED ANTIMICROBIAL THERAPY
General principles of regimen selection — Targeted antimicrobial therapy is chosen based on the results of culture and susceptibility testing from appropriate specimens. Most antibiotic regimens that cover coliforms and anaerobes have comparable efficacy [32-34].
One meta-analysis evaluated 40 randomized or quasi-randomized controlled trials of antibiotic regimens in the treatment of secondary peritonitis in adults [32]. All antibiotics (16 different comparative regimens) showed equivalent clinical success. A subsequent systematic review identified 16 trials that compared various regimens for complicated intra-abdominal infections, including ceftriaxone plus metronidazole, piperacillin-tazobactam, ertapenem, imipenem, meropenem, ceftolozane-tazobactam plus metronidazole, and ceftazidime-avibactam plus metronidazole [34]. Clinical success ranged from 75 to 97 percent and comparators were generally of comparable efficacy.
Assessment of culture data — The source and timing of collection of culture material are critical to the utility of culture results in guiding selection of antibiotics. Specifically, culture of a specimen that is collected prior to starting antibiotics, from a site that should be sterile, is the most informative. In the critically ill patient who requires antibiotic treatment prior to collection of cultures from the site of infection, cultures that are collected early in the course (eg, within a few hours of antibiotic initiation) are more meaningful than those that are collected after longer periods of antibiotic exposure.
Culture of a specimen collected days after starting antibiotics, especially if collected from a chronic drain, is more likely to reflect colonizing bacteria that may have developed resistance to the treatment regimen but are not necessarily causing infection in the patient. For this reason, it is advisable to avoid collecting cultures from chronic drains/fistulae, and results of all such cultures should be assessed carefully for clinical relevance before decisions are made to target results of these cultures with antibiotic therapy [11].
Antibiotic stewardship — In the interests of preserving antibiotic efficacy over time for individual patients and populations, narrowing of antibiotics is advisable once a patient has improved and/or results of reliable cultures are available. Lower-risk patients with community-acquired intra-abdominal infection likely do not warrant alteration of therapy if a satisfactory clinical response to source control and initial therapy occurs, even if unsuspected and untreated pathogens are later reported [11].
Anaerobic coverage — The anaerobic bacterial component of intra-abdominal infections is often not determined but assumed and treated empirically. Coverage for anaerobes is often continued for the duration of the antibiotic course even when anaerobes are not isolated from cultures, particularly if the cultures were obtained only after initiation of antibiotics that are active against anaerobes.
Susceptibility of anaerobic pathogens is rarely known at the time that a decision about the appropriate antibiotic regimen is made since results take a long time, laboratory methods for isolating anaerobes are not well standardized, and activity is usually predictable based on in vitro susceptibility testing from reference laboratories [35], clinical trials, and the site of infection [8,36].
Parenteral versus oral therapy — For patients who are able to eat and tolerate oral medications and whose relevant organisms are not resistant to oral agents, an intravenous regimen can be transitioned to an oral regimen once the patient has demonstrated clinical improvement. Reasonable oral regimens that cover common gut aerobic and anaerobic bacteria include levofloxacin (750 mg once daily) or ciprofloxacin (500 mg twice daily), each with metronidazole (500 mg three times daily), or monotherapy with amoxicillin-clavulanate (875/125 mg two to three times daily), depending on susceptibility testing.
Considerations for specific pathogens
●Enterococcus spp – These are commonly present in intra-abdominal infections and often are not covered in the recommended empiric regimens for community-acquired infections. We agree with the recommendation from the joint Surgical Infection Society and IDSA guidelines that coverage for Enterococcus is not necessary unless it is either recovered from the blood or is the only isolate recovered in culture from the infected site [11].
●Candida spp – Antifungal coverage is warranted if there is growth of Candida spp from a sterile site. Fluconazole is appropriate for Candida albicans; an echinocandin is appropriate for fluconazole-resistant Candida spp and as empiric antifungal coverage in the critically ill patient while awaiting yeast identification and susceptibility testing results. (See "Management of candidemia and invasive candidiasis in adults".)
●Resistant gram-negative bacilli – Isolation of resistant strains of P. aeruginosa, Acinetobacter spp, extended spectrum beta-lactamase (ESBL), or carbapenemase-producing Enterobacteriaceae may warrant specific adjustment of the antibiotic regimen. The novel cephalosporin-beta-lactamase inhibitor combinations, ceftazidime-avibactam and ceftolozane-tazobactam, when combined with metronidazole for anaerobic coverage [37,38], may have niche roles in intra-abdominal infections caused by mixed flora that include P. aeruginosa resistant to other antibiotics or ESBL-producing organisms, as long as susceptibility to these agents is confirmed. Ceftazidime-avibactam has activity against many Klebsiella pneumoniae carbapenemase (KPC)-producing isolates as well. Management of infections due to these organisms is discussed elsewhere. (See "Principles of antimicrobial therapy of Pseudomonas aeruginosa infections" and "Acinetobacter infection: Treatment and prevention" and "Extended-spectrum beta-lactamases", section on 'Treatment options' and "Overview of carbapenemase-producing gram-negative bacilli", section on 'Treatment'.)
●Actinomyces – Actinomyces are slow-growing filamentous gram-positive anaerobic bacteria and are part of the normal flora of the mouth and gastrointestinal tract. If they breach the mucosal surface, these organisms can cause actinomycosis, an uncommon granulomatous disease that is indolent and sometimes mistaken for malignancy because of local spread across tissue planes. In the abdomen, actinomycosis most commonly involves the appendix and ileocecal region. Classic actinomycosis generally warrants prolonged antibiotic therapy (ie, 6 to 12 months). (See "Abdominal actinomycosis".)
In patients who present clinically with a syndrome more typical of an abscess caused by gram-negative organisms and anaerobes (acute presentation, fevers, septic physiology), the isolation of actinomyces on culture of the infected site is of uncertain clinical significance. In such cases, we generally include an antibiotic that is active against actinomyces in the regimen (eg, penicillins) and continue for one to two months, longer than the duration typically warranted for intra-abdominal infections. We also monitor closely for symptoms suggestive of classic actinomycosis with a low threshold for repeat imaging if there is concern for recrudescent infection following antibiotic discontinuation.
Infectious disease consultation — Consultation with an expert in infectious diseases can be especially helpful in the setting of diagnostic uncertainty, for assessment of culture results to guide narrowing of empiric antibiotics, and in specific complex situations. These include neutropenic, organ transplant, or otherwise-immunocompromised patients and patients with antibiotic allergies, potentially infected foreign material (eg, mesh), intra-abdominal malignancy, inflammatory bowel disease, fistulae, or morbid obesity.
Duration of therapy — The appropriate duration of antimicrobial therapy depends on whether the presumptive source of the intra-abdominal infection has been controlled.
When adequate source control has been achieved and the contaminated material cleared from the intra-abdominal space, we generally limit antimicrobial therapy to four to five days [11]. The efficacy of such a short course of antimicrobial therapy was demonstrated in the Study To Optimize Peritoneal Infection Therapy (STOP-IT) trial, in which 518 patients with complicated intra-abdominal infection and adequate source control were randomly assigned to receive either a fixed course of antibiotics for 4±1 days (experimental group) or antibiotics until two days after resolution of fever, leukocytosis, and ileus, with a maximum of 10 days of antimicrobial therapy (control group) [39]. The median duration of antibiotics was four days in the experimental group versus eight days in the control group. The composite primary outcome of surgical site infection, recurrent intra-abdominal infection, or death occurred in a similar percentage of patients in both groups (21.8 in the experimental group versus 22.3 percent in the control group). No significant between-group differences were observed in the individual rates of the components of the primary outcome. Similarly, in a trial of critically ill patients with postoperative intra-abdominal infections and adequate source control, the 45-day mortality rate was not different with a short versus longer antibiotic course (8 versus 15 days) [40].
However, there are several situations in which a longer course of antibiotic therapy is appropriate.
●For patients in whom source control is known to be suboptimal, the optimal duration of antibiotic treatment is uncertain and decisions on treatment duration must be made on a case-by-case basis.
●For those patients with uncomplicated appendicitis who do not undergo immediate surgery, we generally continue antibiotic therapy for approximately 10 days, as in several trials which suggested the safety of this approach in select patients [41]. (See 'Source control and drainage' above.)
●In some cases, an indwelling catheter is required for ongoing drainage and removal of infected material. We generally continue antibiotics in such cases until the efficacy of catheter drainage is established, for example, until an infected hematoma is liquefied well enough to drain effectively via catheter. This may require two to three weeks for a peritoneal abscess. Liver abscess is typically treated for four to six weeks (see "Pyogenic liver abscess", section on 'Treatment'). If there is chronic active drainage through the catheter from an ongoing bowel or biliary leak without accumulation in the peritoneal cavity, discontinuation of antibiotics is usually reasonable as long as the patient has clinically improved.
Patients with undrained abscesses, uncontrolled ongoing bowel leak, or other unresolved mechanical problems often develop worsening clinical signs and symptoms of infection after antibiotics are stopped. We generally refer such patients for repeat surgical or percutaneous intervention for source control. For patients in whom source control cannot be achieved, long-term antibiotics are unlikely to be helpful.
In uncertain cases, declining inflammatory markers (such as C-reactive protein [CRP], erythrocyte sedimentation rate [ESR], and, if available, procalcitonin) can be used cautiously, in addition to clinical resolution, to support antibiotic discontinuation with clinical follow-up to assess for signs or symptoms of recurrent infection. Evidence to support this practice remains indirect.
Several studies have evaluated the utility of inflammatory markers to assist in guiding antibiotic discontinuation, but most of these have been performed with procalcitonin, which is not widely available. In one study of patients with secondary peritonitis who underwent emergency surgery, antibiotic discontinuation based on procalcitonin thresholds (level <1.0 ng/mL or >80 percent decrease compared with the first postoperative day), in addition to resolution of clinical signs, was associated with a shorter duration of antibiotic use and similar adverse events compared with historical controls [42]. In a separate retrospective study, procalcitonin guidance was associated with a 50 percent reduction in antibiotic duration (5 versus 10 days) among intensive care unit patients with secondary peritonitis (both with and without septic shock) [43]. However, in another study of patients with perioperative septic shock in the setting of intra-abdominal infections, the rate and degree of procalcitonin decrease failed to accurately predict treatment response [44].
Evidence from randomized trials supporting antibiotic discontinuation based on procalcitonin levels is mainly from other populations, and these studies have excluded individuals with intra-abdominal abscesses [45]. Additional studies in patients with intra-abdominal infections are warranted.
CLINICAL FAILURE — Patients, particularly those with uncertain source control, should be assessed clinically during antibiotic therapy and after discontinuation for treatment failure, which is suggested by persistent or recurrent signs and/or symptoms of infection, including fever, hypotension, nausea, abdominal pain, organ dysfunction, or leukocytosis. In such cases, the possibility of inadequate source control (eg, an undrained abscess, active bowel leak, retained infected mesh) should be assessed with repeat imaging. The original microbiologic data and the antibiotic regimen should also be reviewed to ensure that the clinically relevant pathogens have been appropriately covered. As noted, most clinical treatment failures are due to failure to achieve source control; cultures from chronic drains, surface wounds, or other nonsterile sites cannot be relied upon to identify organisms requiring targeted antibiotics.
Other considerations in patients with ongoing infectious symptoms or signs include other nosocomial infections (eg, Clostridioides [formerly Clostridium] difficile colitis, health care-associated pneumonia or urinary tract infection, or catheter-associated bloodstream infection). For patients with persistent clinical symptoms and signs but in whom no evidence of a new or persistent infection is uncovered after a careful investigation, discontinuation of antimicrobial therapy is warranted [11]. Noninfectious processes such as thromboembolic disease, drug reaction, and pancreatitis are potential mimickers of infection.
SOCIETY GUIDELINE LINKS — Links to society and government-sponsored guidelines from selected countries and regions around the world are provided separately. (See "Society guideline links: Intra-abdominal infections in adults".)
SUMMARY AND RECOMMENDATIONS
●Pathogenesis – Intra-abdominal infections usually arise after a breach in the normal mucosal defense barrier that allows normal bowel flora to inoculate the abdominal cavity.
●Microbiology – The predominant bacteria involved are coliforms (eg, Escherichia, Klebsiella spp, Proteus spp, Enterobacter spp), streptococci, and anaerobic bacteria. Although enterococci are frequently isolated, their clinical relevance is generally limited to health care-associated infections. (See 'Microbiology' above.)
●Importance of surgical intervention or percutaneous drainage – These procedures are usually critical to the management of intra-abdominal infections (other than spontaneous peritonitis) since most clinical treatment failures are due to failure to achieve source control. Specimens should be obtained during the procedure for microbiologic stain and culture. (See 'Source control and drainage' above.)
The surgical management of intra-abdominal processes is discussed in detail elsewhere. (See "Management of acute appendicitis in adults" and "Acute cholangitis: Clinical manifestations, diagnosis, and management", section on 'Biliary drainage' and "Treatment of acute calculous cholecystitis" and "Acalculous cholecystitis: Clinical manifestations, diagnosis, and management", section on 'Management' and "Overview of gastrointestinal tract perforation", section on 'Indications for abdominal exploration' and "Acute colonic diverticulitis: Surgical management".)
●Antibiotic therapy
•Empiric antibiotic selection – Critically ill patients should receive empiric antimicrobial therapy as soon as possible, ideally once blood and urine cultures have been obtained. For patients who are not critically ill, antibiotic therapy should be delayed until after collection of culture samples from the site of infection.
-Mild to moderate community-acquired infections with no risk factors for antibiotic resistance or poor outcome (table 5) – Antibiotics for these infections should cover enteric streptococci, nonresistant Enterobacteriaceae, and anaerobes (table 2). Examples of infections in this category include perforated appendix or appendiceal abscess. (See 'Low-risk community-acquired infections' above.)
-Severe community-acquired infections or infections in patients at high risk for resistance or poor outcome (table 5) – For these infections, we generally include antibiotic coverage against Pseudomonas aeruginosa, Enterobacteriaceae, enteric streptococci, and anaerobes (table 3). Empiric antifungal coverage is not typically necessary, but is reasonable for critically ill patients with an upper gastrointestinal source. (See 'High-risk community-acquired infections' above.)
-Health care-associated infections – For these infections, we generally include coverage against P. aeruginosa, resistant Enterobacteriaceae, streptococci, enterococci, and anaerobes (table 4). Empiric antifungal coverage is reasonable for patients with upper gastrointestinal perforations, recurrent bowel perforations, surgically treated pancreatitis, heavy colonization with Candida spp, or microbiologic evidence of yeast on intra-abdominal specimens. (See 'Health care-associated infections' above.)
•Targeted antibiotic selection – Depending on the results of culture and susceptibility testing, antibiotic adjustments may be necessary. Regardless of culture results, coverage for Enterobacteriaceae and anaerobes is often continued for the full duration of therapy, even if they did not grow in culture. (See 'Targeted antimicrobial therapy' above.)
Cultures from a chronic indwelling drain more likely reflect colonizing bacteria rather than clinically relevant pathogens. (See 'General principles of regimen selection' above.)
•Duration of antibiotic therapy – For patients who have adequate source control, we suggest limiting antibiotic therapy to four to five days (Grade 2B). Longer courses are often appropriate if source control is suboptimal or uncertain. (See 'Duration of therapy' above.)
