This 9-month, prospective, non-controlled, intervention study comprised two phases: a 3-month pre-intervention phase (Phase I) and a 6-month intervention phase (Phase II), during which the use of third-/forth-generation cephalosporins was restricted and replaced by piperacillin–tazobactam. Rectal swabs were obtained within 24 h after admission (baseline screening), weekly, and 48 h before discharge during Phase I and the last 3 months of Phase II (Phase IIb). Swabs were tested for E. coli and K. pneumoniae, and extended spectrum β-lactamase production was detected with the double disc test.
Use of third/fourth-generation cephalosporins decreased by 63.0 % and 100 % , respectively; while the use of piperacillin–tazobactam increased by 28-fold. The rate of acquisition of extended spectrum β-lactamase-producing E. coli and K. pneumoniae together in rectal swab specimens decreased in Phase IIb as compared with Phase I (19.5 % vs 29.5 % ). Few rectal swab specimens were positive for extended spectrum β-lactamases-producing K. pneumoniae, and no substantial decrease in the rate of its acquisition was observed.