Changing physicians' behavior using combined strategies and an evidence-based protocol.
OBJECTIVES: To review the number of superficial cultures performed in the evaluation of neonatal sepsis and to validate physicians' compliance with an established protocol. DESIGN: Before-and-after-comparison using medical audit. SETTING: Intensive care nursery in a secondary level perinatal facility. PARTICIPANTS: One hundred seventy-five consecutive newborns admitted during a 6-month period for the assessment and potential treatment of sepsis. Eligible patients met the entrance criteria of having a superficial culture performed in conjunction with at least one deep culture, which included cerebrospinal fluid, blood, or urine. This cohort was compared with 205 patients before the introduction of the protocol. INTERVENTIONS: A second audit of physician practice 9 months following the introduction of an evidence-based hospital protocol to discontinue the use of superficial cultures in the diagnostic assessment of neonatal infection. Evidence-based medicine, an opinion leader, continuing medical education rounds, immediate feedback through direct encounters with physicians, and barriers in accessing microbiological tests were used to alter physician behavior. MEASUREMENTS/RESULTS: A significant reduction from 50.5% to 6.9% was achieved in the proportion of superficial cultures performed and a substantial cost savings of $4454.84 was realized without incurring patient morbidity. CONCLUSIONS: A methodologically rigorous reaudit process with planned, interventional strategies may be used as part of a continuous quality improvement program to affect change in physicians' practices. Existing hospital practice standards should be reevaluated against emerging scientific evidence.