Does a patient-level quantitative review of randomized trials on the outcomes in candidemia and invasive candidiasis need to include all patients?

TO THE EDITOR—We have read with interest the patient-level quantitative review of randomized trials for the treatment of invasive candidiasis by Andes et al [1]. However, we do have some questions regarding the figures of included patients. Andes et al report that all modified intent-to-treat (MITT) patients of the 7 analyzed trials were included. Excluded from analysis were cases with missing fungal species information and with multiple Candida species infection, as was one trial arm receiving combination therapy. It is unclear how the data of the sequential treatment arm (amphotericin B followed by fluconazole) in the randomized comparison with voriconazole were managed [2]. In studying the original published trials, a total of 2224 MITT patients are detected, compared to only 1915 patients included in the Andes analysis (Table 1). No explanation is provided as to why 309 patients were excluded from analysis; this is a significantly higher number than would be expected from