Efficacy and tolerability of piperacillin/tazobactam versus ceftazidime inassociation with amikacin for treating nosocomial pneumonia in intensive care patients: a prospective randomized multicenter trial

Objective: To compare clinical and bacteriological efficacy as well as tole rability of two regimens of broad-spectrum antibiotics (ceftazidime versus piperacillin/tazobactam) combined with amikacin in the treatment of nosocom ial pneumonia in intensive care patients. Design: Open label, prospective, multicenter, and randomized phase III clin ical trial. Setting: Medical or surgical intensive care units (ICUs) of nine acute-care teaching hospitals in Spain. Patients and participants: One hundred and twenty-four ICU patients with no socomial pneumonia and requiring mechanical ventilation were included. They were randomized to receive amikacin (15 mg/day divided into two doses) com bined with either piperacillin (4 g every 6 h) and tazobactam (0.5 g every 6 h) (n = 88) or ceftazidime (2 g every 8 h) (n = 36). Measurements and results: The causative pathogen was determined in 60.2 % o f patients in the group of amikacin plus piperacillin/tazobactam and in 76. 9 % in the group of amikacin plus ceftazidime. A total of 94 bacterial orga nisms were isolated among which gram-negative bacilli predominated. Pseudom onas aeruginosa being the most frequent. Clinical response at the end of an tibiotic therapy was considered satisfactory (cure and/or improvement) in 6 3.9 % of patients in the amikacin plus piperacillin/tazobactam group and in 61.5 % in the amikacin plus ceftazidime (odds ratio 1.1; 95 % confidence i nterval 0.44-2.75). Eradication or presumptive eradication rates for each p athogen and for either gram-negative or gram-positive bacteria were similar in both antibiotic combinations (odds ratio 1.2; 95 % confidence interval 0.39-3.66). A total of 21 adverse effects (23.9 %) occurred in the amikacin plus piperacillin and tazobactam group and six (16.7 %) in the amikacin pl us ceftazidime group, thrombocytosis, renal dysfunction, and hepatic cytoly sis being the most common. The efficacy and tolerability of the two therape utic regimens were similar not only in the whole study population, but also in the subset of p. aeruginosa-related pneumonia (odds ratio 1; 95 % confi dence interval 0.08-13.37). Conclusions: Amikacin associated with either ceftazidime or piperacillin an d tazobactam has shown comparable efficacy and tolerability in the treatmen t of ICU patients with nosocomial pneumonia.