Published data have suggested a correlation between antifungal prophylaxis and bacteremia in febrile neutropenia. This correlation was investigated among 3002 febrile neutropenic patients enrolled in 4 trials during 1986-1994. Globally, 1322 patients (44%) did not receive antifungal prophylaxis; 835 (28%) received poorly absorbable antifungal agents and 845 (28%) received absorbable antifungal agents. The rates of bacteremia for these groups were 20%, 26%, and 27%, respectively (P = .0001). In a multivariate model without including antifungal prophylaxis, factors associated wtih bacteremia were: age, duration of hospitalization, duration of neutropenia before enrollment, underlying disease, presence of an intravenous catheter, shock, antibacterial prophylaxis, temperature, and granulocyte count at onset of fever. When antifungal prophylaxis was included, the adjustment quality of the model improved slightly (P = .05), with an odds ratio of 1.19 (95% confidence interval [CI], 0.92-1.55) for patients receiving nonabsorbable and 1.42 (95% CI, 1.07-1.88) for those who were receiving absorbable antifungal agents. Antifungal prophylaxis with absorbable agents might have an impact on the rate of documented bacteremia in febrile neutropenia. This effect should be confirmed prospectively.
ASJC Scopus subject areas