Early newborn discharge: a tale of two hospitals.

This study reviews physician documentation of compliance with The American Academy of Pediatrics (AAP) policy RE9539 regarding early newborn hospital discharge and follow-up. All pediatricians in Kalamazoo, MI, were educated at a grand rounds regarding the AAP early newborn discharge policy. Newborns are seen at 2 community hospitals. One of them simultaneously instituted a Quality Improvement/Feedback (QI/F) program regarding early newborn discharge, (intervention) and the other 1 did not (control). This is a retrospective chart analysis that compares performance of 4 pediatric practices seeing newborns at each hospital. Each practice was compared at each hospital for appropriateness of discharge orders before and after the educational grand rounds and the QI/F initiative. Statistical analysis was done using the chi square test and the Breslow-Day test for homogeneity, and the Fisher's Exact Test. Odds ratios with a 95% confidence interval based on Taylor's approximation were used. There were no significant differences between the pediatric practices' performance before and after the educational initiative at the control hospital. There were significant differences before and after the educational initiative in the intervention hospital with the QI/F initiative. There was a significant reduction in variation among the practices after the QI/F initiative at the intervention hospital. When both hospitals were compared after the educational initiative, there was a significant difference between compliance among the same practices at each hospital, with better compliance at the intervention hospital with the QI/F initiative. There were significant differences in physicians' performance at the intervention hospital before and after the educational and QI/F initiatives. However, it was noted that the very same physicians did not comply as well in the control hospital without the QI/F initiative, thus still raising questions as to whether QI measures alter physician "culture." It is possible that the driving force for change in physician behavior was more intragroup peer pressure than an external QI/F initiative.