Hepatic abscess: sensitivity of imaging tests and clinical findings.

The imaging studies of 63 patients with hepatic abscess were evaluated to determine the sensitivity of specific imaging tests and define causes of false-negative test results. Computed tomography (CT) detected 57 of 59 (97%) separate episodes of hepatic abscess. The two false-negative CT examinations were in patients with a diffuse low-density pattern throughout the liver. The radionuclide (RN) examination detected 16 of 20 (80%) cases, missing abscesses less than 2 cm in diameter. Ultrasound detected 33 of 42 (79%) cases, missing abscesses in the dome of the liver, small abscesses, and 2 large early abscesses. For all three imaging modalities, a specific diagnosis of abscess was possible only in those patients in whom CT scans demonstrated abscess gas (15%). The results of the imaging studies were correlated with the patients' clinical condition and laboratory findings. Thirty-one percent of patients were afebrile while 23% had normal white blood cell counts. Thirteen percent had totally normal liver function tests. We conclude that the clinical absence of fever, leukocytosis, or elevated liver function tests does not permit exclusion of the diagnosis of hepatic abscess. A CT scan is highly specific in excluding the diagnosis of hepatic abscess in the absence of diffuse liver disease. In this retrospective study CT was the most sensitive imaging modality available for the detection of hepatic abscess.