Data Collection
The potential case list obtained from HIM was first reviewed to remove duplicate cases, confirm cancer diagnosis inclusion criteria (ALL, AML, or lymphoma), and exclude any patients with underlying immunodeficiency or who developed IMI after HSCT or organ transplant. For the remaining patients, demographic, laboratory, clinical and radiographic data were manually abstracted from the electronic medical record system using a standardized, comprehensive data abstraction form. Demographic and clinical characteristics of the total number of cases diagnosed during the study time period were obtained from records maintained by the clinical disease-based teams and verified in the medical record.
All remaining potential IMI cases were then classified as meeting or not meeting criteria for probable or proven IMI, as defined by the 2008 European Organization for Research and Treatment of Cancer/Mycoses Study Group (EORTC/MSG) Consensus Group.9 To make this determination, records were reviewed by two independent members of the research team (AD, JB, KR, PH, AM, AI, MC), and discordant entries were resolved through group discussion. All mold isolates noted in the medical record were identified according to standard laboratory practices. The date of diagnosis of IMI was defined as the date of first positive culture and/or diagnostic study. Clinical information regarding the underlying disease status (date of cancer diagnosis, type of cancer, and whether the patient was receiving upfront treatment or treatment for relapsed disease), relevant exposures concurrent to the diagnosis of IMI, and relevant host factors were abstracted for all patients. Factors previously established as risk factors for IMI were noted if present within 2 weeks of the IMI diagnosis, such as neutropenia (at least one incidence of absolute neutrophil count <500 cells/mm3), hyperglycemia (at least one incidence of random blood glucose level >200 mg/dl), intensive care unit admission, and concomitant bacterial infection(s) requiring antibiotics.