Adaptation of clinical guidelines to develop a risk-based algorithm for antifungal prophylaxis
Development of an evidence-based algorithm for antifungal prophylaxis, including broader coverage prophylaxis, was initially prompted by a relatively high incidence of IMI noted in specific leukemia patient populations treated prior to 2016 compared with reported frequencies of IMI in the published literature. Prior to 2016, routine antifungal prophylaxis was employed for a limited population of children diagnosed with high-risk hematologic malignancies. For example, all children diagnosed with AML were prescribed fluconazole, but there was inconsistent practice for prescribing prophylaxis to children with relapsed disease or in other high risk groups. If antifungal prophylaxis was strongly recommended on a particular clinical trial, such prophylaxis was given to all patients treated according to that regimen, regardless of their enrollment to the research study. In 2015, we established a multidisciplinary team of TCH Cancer and Hematology Center Leukemia team providers, clinical pharmacy specialists, and members of the TCH Infectious Diseases team to identify and address factors that may be contributing to excess IMI in children with hematologic malignancies. The team reviewed incident cases in patients with oncologic diagnoses, and hospital-wide patterns of mycoses epidemiology. Available evidence-based guidelines were reviewed, considering specific at-risk populations and perceived intensity of chemotherapy treatment, prior to reaching a consensus for a risk-based algorithm and flow chart (Figure 1 ). This clinical resource was finalized and implemented in January of 2016, an effort that included delivery of section-wide education to TCH Cancer and Hematology Center faculty and fellows.
Change in IMI frequency pre and post-implementation of an antifungal risk-based algorithm
Implementation of an evidence-based, risk-adapted algorithm for antifungal prophylaxis reduced the frequency of IMI in patients with hematologic malignancies by 40%, from 4.8% to 2.9%. The decrease in IMI incidence was most pronounced among patients with AML, likely due to a change in practice that broadened antifungal coverage to include molds (Table 2 ).