- Piperacillin/tazobactam (P/T) has been used in NICU since 1999 as part of the empirical treatment for presumed sepsis. We studied the microbiological and clinical efficacy and safety of P/T use in newborns with sepsis, using a retrospective analysis of medical records of all newborns treated with P/T (in association with amikacin) during 1999-2003. P/T plus amikacin was used for 353 episodes of presumed sepsis occurring in 252 newborns: 105 episodes occurred in 100 newborns treated for 5-14 d while in 248 (70%) episodes the treatment was discontinued after 2-3 d. 123 pathogens were isolated from the blood cultures (40 in the 5-14 d group and 83 in those treated for 2-3 d only). There were 56 Gram-negative, 55 Gram-positive and 12 Candida spp. isolates. Klebsiella spp. (40%) and E. coli (30%) were the most commonly isolated Gram-negative pathogens; coagulase-negative Staphylococcus (CONS) represented 95% of the Gram-positive pathogens isolated. Pathogen eradication was achieved within 48-72 h of therapy in 93% Gram-negative organisms. There were no clinical, laboratory or cranial ultrasound adverse effects associated with P/T use. We conclude that P/T in association with amikacin is microbiologically and clinically efficacious and safe in the treatment of sepsis in newborns.