laparoscopic surgery versus laparotomy for this indication (43). Percutaneous drainage has a
low success rate of just 32% and is generally insufficient management except in the case of a
well-defined abscess, or one remote from the pancreas (41). Runzi et al. recently published a
small study in which antimicrobial therapy alone resulted in similar outcomes to
antimicrobials combined with surgery (42); however, nonsurgical management is not currently
common practice for infectious necrotic pancreatitis.
Abdominal compartment syndrome has been noted in severe acute pancreatitis and
decompression has been suggested for patients whose transvesical intra-abdominal pressure
reaches 10 to 12 mm Hg (43).
An appropriate antibiotic regimen for infected pancreatic necrosis is the second arm of a
successful treatment plan: given the wide range of possible offending organisms, a Gram stain
is recommended to tailor specific initial therapies prior to culture results. For gram-negative
organisms, a single-agent carbapenem is effective; for gram-positivesb-lactamase–resistant
drugs, vancomycin, and even linezolid must considered. When yeast is identified, high-dose
fluconazole or caspofungin should be sufficient. In any case, if infection develops despite
antibiotic prophylaxis, a different class of drugs must be administered for treatment than was
given for prophylaxis (44).
A meta-analysis by Bassi et al. found that antimicrobial prophylaxis for patients with
necrotic pancreatitis successfully decreases the incidence of infection by half and triples overall
survival (45). Although current literature does not specifically favor any specific antibiotic as
prophylaxis, it is nonetheless clear that microbial coverage must be broadly targeted. One- to
two-week courses of cefuroxime, imipenem with cilastin, and ofloxacin with metronidazole
have each been tried with success (42).
MIMICS OF ABDOMINAL INFECTION
Multiple conditions may mimic a postsurgical abdominal infection and must be considered
when searching for diagnosis. An exhaustive list of these is beyond the scope of this chapter;
however, the reader should be aware of the general possibilities. Fever, for instance, in the
postoperative patient, is not always secondary to infection. Particularly relevant to the
postsurgical patient are events such as atelectasis, myocardial infarction, stroke, hematoma
formation, and even pulmonary embolism that may occasionally present with a fever
component. Other causes that warrant deliberation include drug or transfusion reaction,
malignancy, collagen vascular disease, endocrine causes such as hyperthyroidism, and less
common etiologies such as disordered heat homeostasis secondary to an ischemic
hypothalamic injury or even familial malignant hyperthermia. Pain is yet another symptom
that may be misleading: Abadir et al. published a study in which patients with segmental
infarction of the omentum or epiploic appendages presented with localized peritonitis,
mimicking appendicitis, diverticulitis, and cholecystitis (46). Furthermore, it is important to
interpret radiological findings with an open mind. A fluid collection on CT does not
necessarily represent an abscess. Again, high on the differential that must be considered is
hematoma, and one may explore other diagnoses given the individual patient history. For
example, Yu et al. found that the fundus of the excluded stomach following gastric bypass
surgery may fill with air, fluid, and contrast material, thus closely resembling a loculated fluid
collection (47). Finally, entertain where appropriate the idea of extra-abdominal infections. A
myocardial infarction involving the inferior wall of the heart and lower lobe pneumonias, for
instance, may present with abdominal pain and fever despite extra-abdominal origins.
BLOODSTREAM INFECTION
Bloodstream infection, defined as a positive blood culture with organisms of intra-abdominal
origin, is associated with mortality just over 60% (48). Gram-negative organisms, andE. coliin
particular, are most common. Approximately 40% of all organisms isolated by DeWaele and
colleagues at Ghent University hospital were multidrug resistant. Methicillin-resistant
Staphylococcus aureusand extended-spectrumb-lactamase–producing (ESBL) organisms are a
growing concern. Each ICU will likely have a unique pattern of pathogens with differing
antimicrobial susceptibilities; therefore, the clinician should be up-to-date on current
antibiograms for resistant flora in the critical care unit.
268 Wilson