[Skip to Content]
[Skip to Content Landing]
Citations 0
June 7, 2000

Improving the Conduct and Reporting of Clinical Trials

Author Affiliations

Phil B.FontanarosaMD, Deputy EditorIndividualAuthorStephen J.LurieMD, PhD, Contributing EditorIndividualAuthor

JAMA. 2000;283(21):2787-2790. doi:10.1001/jama.283.21.2787

To the Editor: Drs Johansen and Gøtzsche1 report on difficulties they experienced in conducting a meta-analysis of clinical studies on fluconazole and discuss a potential bias introduced by the methods of these studies. The authors criticize studies comparing oral fluconazole vs oral amphotericin B for antifungal prophylaxis in patients with cancer. In their meta-analysis, they included studies that used nystatin to reduce Candida colonization (an established risk factor for invasive mycoses) in an attempt to prevent fungal infections that might result from this colonization. Nystatin was used as a comparator only in this context but has been shown to be effective against oropharyngeal and esophageal candidiasis.2 The use of nystatin was additionally justified because of superior compliance rates, especially in children. However, nystatin was never used to counter systemic fungal infections directly, because it is known to be ineffective systemically due to poor intestinal absorption. The criticism of a 3-armed study being reported as a 2-group design (fluconazole vs polyenes) is justified.