It is advocated that prolonged infusion of beta-lactamic antibiotics provides better bactericidal effect. The aim of the present study is to randomize patients a to extended cefepime infusion regimen (lasting four hours) or to a usual infusion regimen (not lasting more than thirty minutes) and evaluate the clinical efficacy of this theoretical pharmacokinetic advantage.

Introduction: it has been proved that the main determinant for microbiological cure gram-negative infections treated with cefepime is the time above minimal inhibitory concentration (T>MIC). Although this particular pharmacokinetic property have never been proved in clinical trials utilizing cefepime, intervention studies with other betalactamic antibiotics, such as pipercillin-tazobactam, showed clinical benefit.

Objective: to evaluate if prolonged infusion (lasting 4 hours) with cefepime translate into better clinical outcomes.

Methods: the investigators aim to conduct an open-label, unique-centered, randomized controlled trial using cefepime in prolonged infusions in patients being treated for urinary or respiratory tract infections. Patients developing these infections after 72 hours of hospital admission, requiring the use of broader spectrum antibiotics after clinical failure or isolating gram-negative bacteria from adequate sample sensible to cefepime will be enrolled. The use of a second antibiotic such as clindamycin, vancomycin or metronidazole will be allowed.

Interventions: the investigators aim to randomize 134 patients to two different treatment arms, the intervention arm who will receive the medication in a four hour lasting infusion, and the active control arm who will receive the medication in a 30 minutes lasting infusion. The randomization will be conducted in blocks of ten patients each, and it will be balance according to the patients age (older or younger than 65 years old) and presence of SIRS criteria.

Inclusion Criteria:

• Urinary tract infection after 48-72 hours of hospital stay
• Respiratory tract infection after 48-72 hours of hospital stay
• Catheter related urinary tract infection after 48-72 hours of hospital stay
• Urinary or respiratory infections not responding to a narrower spectrum antibiotic
• Isolation of bacteria sensible to cefepime in blood, urine or sputum in a clinical context of infection

Exclusion Criteria:

• Glomerular filtration rate lower than 30 ml.min/1,73m²
• ICU admission for at least 72 hours before randomization
• Neutrophil count lower than 1000 cels per mm³
• Hematologic malignancy, bronchiectasis and cystic fibrosis
• Patients allergic to cefepime
• Concomitant treatment to another infectious disease
• Central nervous system, cutaneous or intrabdominal infections
• Solid organ transplantation
• HIV with a cluster differentiation 4 (CD4) lower than 100 cels per mm³