Significant reduction of nosocomial sepsis rates at two neonatal intensive care units

Background. Nosocomial infection is one of the most prevalent sources of concern in neonatal intensive care units (NICUs). Significant center-to-center variation has been noted suggesting that site of care influences outcome.

Objective. To reduce the nosocomial sepsis rate by implementing a multiple-approach prevention programme elaborated by a panel of doctors and nurses including microbiologists, infectious disease -, infection control -, and neonatal intensive care specialists.

Methods. Multiple changes in routine management of intensive care patients were introduced in the two NICUs of the tertiary care University Hospital Vienna, Austria. Processes implemented included among others: use of sterile gloves for insertion of periperal venous lines, an educational campaign for central venous catheter insertion and on device use and care, use of sterile gloves for any manipulation on central venous catheters, education of visitors and attendants (i.e. radiologists, surgeons) on hand washing techniques including distribution of a broschure, implementation of a standardized protocol for cleaning of toys in the incubator, introduction of infection-focused ward rounds including dictation of type and duration of antibiotic therapy, repeated education, observation and motivation of NICU staff members by hospital control nurses. The primary outcome measure was episodes of nosocomial sepsis defined as positive blood culture, collected more than 5 days after birth, per 1000 patient days. The secondary outcome measure was the overall consumption of antibiotics per patient day. A “preimplementation“ period of 9 ½ months prior to April 1st 2004 was compared with a 9 ½ months “postimplementation“ period.

Results. In the baseline period, the infection rate was 12.9 episodes per 1000 patient days. In the postimplementation period, the corresponding rate was 9.2 per 1000 patient days. Average total costs per patient day for vancomycin and gentamicin, the first choice empiric antibiotic regimen for nosocomial sepsis, were reduced by more than 30%. None of the improvements reached statistical significance because of sample size.

Conclusion. We conclude that there was a significant reduction in nosocomial sepsis rates. The new measures introduced may have substantially contributed to this reduction.