RESUMO
BACKGROUND/AIMS: In this study, the sensitivity-specificity of galactomannan-enzyme immunoassay (GM-EIA) with a cut-off value of 0.5 for a single, two, or three consecutive positivity in the diagnosis of invasive pulmonary aspergillosis (IPA) in neutropenic patients with hematological malignancy was investigated. METHODS: IPA was classified as "proven," "probable," or "possible" as described in the guidelines prepared by the European Organization for Research and Treatment of Cancer and Mycoses Study Group." Serum samples were collected from the patients twice a week throughout their hospitalization. A total of 1,385 serum samples, with an average of 8.3 samples per episode, were examined. RESULTS: Based on the 165 febrile episodes in 106 patients, 80 (48.5%) were classified as IPA (4 proven, 11 probable, 65 possible) and 85 (51.5%) as non-IPA. The sensitivity/ specificity was 100%/27.1% for a single proven/probable IPA with the cut of value of GM-EIA > or = 0.5, 86.7%/71.8% for two consecutive positive results, and 73.3%/85.9% for three consecutive positive results. CONCLUSIONS: With the galactomannan levels measured twice a week, consecutive sensitivity decreased and specificity increased. Therefore, an increase may be obtained in sensitivity-specificity by more frequent monitoring of GM-EIA starting from the first day of positivity is detected.