Background: Nystatin is sometimes used prophylactically in patients with severe immunodeficiency or in the treatment of fungal infection in such patients, although the effect seems to be equivocal.

Objectives: To study whether nystatin decreases morbidity and mortality when given prophylactically or therapeutically to patients with severe immunodeficiency.

Search strategy: MEDLINE and The Cochrane Library using a comprehensive search strategy, date of last search November 2001. Contacted industry and scanned reference lists.

Selection criteria: Randomised trials comparing nystatin with placebo, an untreated control group, fluconazole or amphotericin B.

Data collection and analysis: Data on mortality, invasive fungal infection and colonisation were extracted by both authors independently. The outcomes were weighted by the inverse variance. A random effects model was used unless p>0.10 for the test of heterogeneity.

Main results: We included 12 trials (1,464 patients). The drugs were given prophylactically in ten trials and as treatment in two. Seven trials were in acute leukaemia, two in cancer, one in liver transplant patients, one in critically ill surgical and trauma patients, and one in AIDS patients. Nystatin had been compared with placebo in three trials and with fluconazole in nine; the dose varied from 1.5 MIE to 72 MIE daily. The effect of nystatin was similar to that of placebo on fungal colonisation (relative risk 0.85, 95% confidence interval 0.65 to 1.13). There was no statistically significant difference between fluconazole and nystatin on mortality (relative risk 0.76, 0.49 to 1.18) whereas fluconazole was more effective in preventing invasive fungal infection (relative risk 0.37, 0.15 to 0.91) and colonisation (relative risk 0.49, 0.34 to 0.70). The results were very similar if the three studies which were not performed in cancer patients were excluded.

Reviewer's conclusions: Nystatin cannot be recommended for prophylaxis or treatment of Candida infections in immunodepressed patients.