Retrospective review of gynecologic oncology patients with therapy-induced neutropenic fever.

We reviewed the course and outcome of gynecologic oncology patients with neutropenic fever (NF), and identified low-risk patients who might be candidates for outpatient management. Charts from patients with the discharge diagnosis of NF from 1990 to 1994 were reviewed for variables related to the febrile neutropenic episode. Outcome was evaluated in terms of the duration of neutropenia, length of hospital stay, NF treatment result, incidence of positive blood cultures, dose reduction in the subsequent course of chemotherapy, and death. Statistical associations between variables and outcome parameters were examined by the Student t test and chi 2 or Fisher's exact tests as indicated. Multivariate analysis by logistic regression was done to determine independent significance of variables. Forty-five episodes of NF were identified involving 40 patients. The median duration of neutropenia following the diagnosis of NF was 2.5 days. The source of fever was unexplained by exam or cultures in 25/45 (56%) episodes. There were two (4%) deaths. In 16/45 episodes, patients had been treated with at least one prior regimen of chemotherapy (median, eight courses). Episodes of NF in patients receiving second-line chemotherapy were associated with a prolonged time of neutropenia (> 3 days, P = 0.058); however, this did not translate into increased hospital stay, treatment failure, or death. Thirteen of 45 (29%) NF episodes developed in patients while already hospitalized for medical or surgical conditions. In this group, cultures were positive in 64% of cases. The remaining 32 NF episodes developed in patients while at home. Patients who developed outpatient NF (ONF) had positive cultures in 23% of cases (P = 0.08) and had a median hospital stay of 4 days. No patient with ONF who remained hemodynamically stable during the first 12 hr of admission suffered serious morbidity. In a multivariate analysis, only bacteremia approached statistical significance in predicting a longer hospital stay (P = 0.07), and no variable studied was predictive of prolonged NF in the ONF group. Our retrospective analysis indicates that patients with ONF who remain stable during the initial 12 hr of hospitalization might safely be discharged home with appropriate antibiotics. Prospective study of outpatient management of NF is required to confirm these findings.