Aim: To evaluate the effect of a change in antibiotic protocol on pathogens that cause neonatal sepsis. Methods: Suspected sepsis was treated with amikacin together with ceftazidime in 1995–1998 and piperacillin/tazobactam in 1999–2002. Results: The annual rate for Klebsiella sepsis fell from 2.5 to 0.45 cases per 1000 admission days (p= 0.0001) between the two periods studied. Conclusion: The change from ceftazidime to piperacillin/tazobactam is associated with a decrease in the incidence of Klebsiella sepsis.

The change from ceftazidime to piperacillin/tazobactam is associated with a decrease in the incidence of Klebsiella sepsis.