This is a retrospective ease-control study in a 499-bed medical center with an AIDS-designated unit of 51 beds and additional AIDS patients on other floors. In the period covered by this survey, we had more than 100 HIV-positive patients with definitive Л/ tuberculosis disease.s Thus, missed tuberculosis that could account for the clinical symptoms is unlikely. The computer-based research was performed on MEDLINE and on additional databases in all languages.
Hospital microbiolog) files were reviewed to identify all isolates of M gordotuw between January 1, 1989 and April 30, 1992. We reevaluated the medical records of all patients who had Л/ gordonae isolated in urine, blood, stool, synovial fluid, bone marrow, duodenal aspirate, sputum, bronchial, and tracheal secretions.
All patients were admitted to a midtown Manhattan teaching hospital and 24 patients had been followed up by at least one of us. Most of the HIV-positive patients had an extensive workup for fever, including repeated cultures of blood (centrifugation lysis system), sputum, urine, stool, cerebrospinal fluid, chest radiographs (CXR), and additional imaging techniques such as sonograms, gallium scans, and computed tomographic (CT) scans. Invasive procedures such as bone marrow aspiration and biopsy, synovial and liver biopsies, and fiberoptic endoscopies were performed w hen clinically indicated. The M gordonae-positive specimens of 17 patients had been sent within the first 2 days of hospitalization. Routine epidemiologic and microbiologic hospital surveillance studies did not reveal contamination of tap water, laboratory solutions, or instrumentation.