Empirical and subsequent use of antibacterial agents in the febrile neutropenic patient.

Unlabelled: The objective of this analysis were an assessment of the feasibility of a more individually tailored approach of empirical antibiotic therapy in febrile neutropenia and an exploration of the reasons to modify the initial regimen.

Design, setting and subjects: The main source was a database on febrile neutropenic cancer patients from an unblinded large trial conducted in 35 centres world-wide. This was supplemented by data from patients enrolled in a consecutive series of randomized trials at the Department of Haematology, University Hospital Nijmegen.

Interventions: Diagnostic procedures were standardized, types of possible infections defined and the reasons for modifying an empirical regimen were recorded.

Main outcome measures: Survival of the febrile neutropenic episode, development of microbiologically and clinically defined infection in relation to causative organisms, and results of modification.

Results: Monotherapy was as effective as combination therapy with an overall mortality of < or = 7%, with 21% of neutropenic episodes accompanied by a clinically defined infection proving fatal compared with only 4% of episodes without a focus. At the end of treatment the empirical regimen had been added to in 60% of cases in the multicentre trial, in contrast to 39% in our own institution, in many cases simply because of continuing fever.

Conclusion: The development of local guidelines for individually tailoring antibiotic therapy by complementing the empirical regimen is a feasible option for achieving an optimal anti-infective strategy for febrile neutropenic cancer patients.