Antibiotic strategy after the empiric phase in patients treated for a hematological malignancy

Empiric broad-spectrum antibiotic therapy has become a generally accepted strategy in the treatment of febrile neutropenic patients. Particularly in patients with prolonged neutropenia, subsequent adaptation of such a regimen will be the rule rather than exception. Since there are no uniformly accepted guidelines for the modification of antibiotic therapy during the post-empiric phase, we assessed the impact of a set of rules that evolved during the first randomized trials. Evaluation of the clinician's compliance with these rules in 1951 febrile neutropenic episodes was the subject of the present analysis. Treatment was modified in 761 (39%) cases, and these changes were made according to the rules in 76%. For 75% of the alterations in treatment during the evening and night shifts, no reasonable explanation was established, while 93% of the modifications during the normal working hours were made for objective reasons. The empiric regimen was more frequently changed in patients with a clinical focus of infection at the onset of fever than in patients who showed fever as the only symptom of a possible infection. The perceived need for modification amounted to 69% in pulmonary infections, to 51% in skin and soft-tissue infections, to 44% in patients with abdominal complaints, and to 37% in upper respiratory tract infections. Glycopeptides constituted 22% of modifications, particularly in patients with a central venous catheter, and systemically active antifungals were administered in 16% of cases. Especially inexperienced clinicians tend to adjust antibiotic therapy, in spite of the fact that persistence of fever alone seldom reflects inadequate treatment when the clinical condition of the patient is stable or improving. On the other hand, the development of subsequent infectious events emphasizes that a genuine need for modification does frequently exist.