ABSTRACT
PURPOSE: Since febrile neutropenic patients were recognized to constitute a heterogeneous population, several models have been developed for predicting the risk of serious medical complications. The Multinational Association for Supportive Care in Cancer score and its derived clinical prediction rules have been validated, but thus far there were no data about its use for simplifying therapy in predicted low-risk patients. PATIENTS AND METHODS: In a single institution, we followed all episodes of febrile neutropenia between January 1999 and November 2003. Those patients predicted at low risk for complications, who were not receiving antibacterials at fever onset and were eligible for treatment with oral antibiotics, were treated with ciprofloxacin and amoxicillin-clavulanate and were discharged if they were clinically stable or improving after an initial observation period. The primary end point of the study was the rate of resolution of the febrile neutropenic episode without complications, among these early discharged patients. RESULTS: Of 383 first febrile neutropenic episodes predicted at low risk of complication, 178 patients (33 men and 145 women, mainly with solid tumors) were treated orally; they constituted the basis of our analysis. Seventy-nine patients (44%) were discharged early (with a median time to discharge of 26 hours); no complications occurred among them but three patients had to be readmitted, resulting in a success rate of 96% (95% CI, 92% to 100%). CONCLUSION: Our study shows that oral therapy followed by early discharge was feasible in a small but significant proportion of patients selected by a strategy combining predicted low risk and medical and nonmedical criteria.