Objective: To compare piperacillin and tobramycin with ceftazidime alone for the empiric treatment of fever in the neutropenic patient without evidence of skin infections or anaerobic infections.
Design: A multicenter, randomized, controlled trial.
Patients: 876 febrile, neutropenic episodes in 696 patients (83% acute leukemia or bone marrow transplantation); 92 episodes were excluded from analysis because of protocol violation.
Interventions: Patients received either intravenous ceftazidime (2 g every 8 h) or piperacillin (12 to 18 g/d in 4 to 6 divided doses) plus tobramycin (1.7 to 2.0 mg/kg body weight every 8 h). Treatment could be modified at any time at the discretion of the investigator.
Measurements: Percentage of satisfactory response, eradication of the infecting organism, development of superinfections, and occurrence of adverse events.
Results: As a single agent, ceftazidime was as effective as the combination of piperacillin and tobramycin (62.7% satisfactory responses compared with 61.1%; odds ratio, 1.07; 95% CI, 0.79 to 1.44; P > 0.2). Equivalent responses were also obtained in episodes of profound neutropenia (odds ratio, 0.76; CI, 0.43 to 1.33; P > 0.2). Infectious mortality was 6% for ceftazidime and 8% for the combination therapy. Eradication of the infecting organisms was achieved in 79% of bacteremic episodes treated with ceftazidime compared with 68% of the episodes treated with the combination therapy (odds ratio, 1.76; CI, 0.92 to 3.38; P = 0.08), and rates for gram-negative rod bacteremia were also similar (95% compared with 77%; odds ratio, 5.25; CI, 1.0 to 27.5; P = 0.03). Superinfections developed in 38 episodes in each group. An adverse event occurred in 8% of episodes treated with ceftazidime compared with 20% of episodes treated with combination therapy (P < 0.001).
Conclusion: Ceftazidime alone was as effective but safer than the combination of piperacillin and tobramycin for the empiric treatment of febrile, neutropenic patients, even those with profound and prolonged granulocytopenia.