Background: Acquired infection is one of the most prevalent sources of concern in neonatal intensive care units (NICUs). Center-to-center variation has been noted by both the National Nosocomial Infection Surveillance System and the Vermont Oxford Network suggesting that site of care influences outcomes including acquired infection.
Objective: To reduce the acquired infection rate by isolating and then implementing meaningful process differences between high and low infection rate centers.
Design/method: A multistaged observation and intervention study. The primary outcome measure was defined as a positive blood culture, collected more than 3 days after birth. Hospital patient days along with infection episodes were collected for all NICU admissions in the network during the baseline and post-implementation periods. A detailed observation guide was used during site visits to high and low infection rate centers. The observations recorded in the guide allowed the team to isolate meaningful differences, which were shared with the network. Individual NICUs decided which of the meaningful differences, if any, to implement. To estimate the impact on costs, additional data were gathered in a case-matched series of infants in one demonstration site.
Results: In all, 15 meaningful differences were isolated and shared with the network. The network rate for acquired infection dropped from 3.8 to 2.9 episodes per 1000 patient days. In the demonstration site, the infection rate dropped from 7.4 to 4.0 per 1000 patient days.
Conclusion: Isolation of process level differences between high and low performing centers followed by implementation of these meaningful differences may reduce acquired infections. Other targeted areas of care may benefit from this quality improvement methodology.