Article type
Year
Abstract
Abstract:To evaluate an opinion-leader strategy to implement evidence-based guidelines based on the Cochrane Collaboration and other systematic reviews, to improve breast-feeding rates under average conditions of practice.
Methods: We matched 18 hospitals in a diverse perinatal region in upstate New York, and randomized them between intervention and control groups. The intervention was an opinion-leader strategy based on patterns of social influence observed in empirical studies of innovation. We surveyed the obstetric clinicians--obstetricians, family practitioners, and nurse midwives--in each hospital to document their knowledge, attitudes, and beliefs, and identify their opinion leaders. Using triangulation, we also interviewed obstetric nurse managers in each hospital. During the study year, 12/96 through 11/97, we recruited the 9 opinion leaders, convened two workshops to assist development of their implementation strategies, provided written materials for clinicians and mothers, and monitored activities at the sites. Using multiple regression, the analysis compared the difference in breast-feeding rates between the prior baseline year and the study year in intervention vs. control hospitals. Besides the intervention group, matched hospital group, and study year, the analysis included independent variables to control for characteristics of mothers and clinicians likely to influence breast-feeding rates. A GEE (generalized estimating equation) procedure adjusted for clustering of mothers within clinician and clinicians within hospitals.
Results: Clinicians responding to the survey (59%) had a high degree of knowledge about the value of breast-feeding and management of associated problems. At each hospital, the 3 obstetric specialities agreed on their identification of the opinion leader, and the nurse manager agreed with the clinicians. All the opinion leaders also held administrative positions. The opinion leaders developed and implemented activities that included department meetings, working with nursing staff and pediatricians, changes in hospital policies, and earlier discussions with mothers. In the multivariate analysis, breast-feeding rates in the intervention hospitals, however, did not differ significantly from those in the control hospitals. This pattern pertained to mothers' intentions to breast feed exclusively and to their intentions to combine breast and formula feeding. Aside from the variables related to the intervention's effectiveness, the independent variables on mothers' and clinicians' characteristics were significantly associated with mothers' breast-feeding intentions.
Conclusions: The project demonstrated the feasibility of identifying common opinion leaders among multidisciplinary health professionals. Relying on nurse managers to identify opinion leaders and the overlap between opinion leaders and administrators merit further exploration. Though not significant, the results help to improve understanding of the circumstances in which different behavior-change strategies are effective. The opinion-leader strategy may be more effective in changing professional behavior than in changing organizational policies, community attitudes, or consumer behavior.
Methods: We matched 18 hospitals in a diverse perinatal region in upstate New York, and randomized them between intervention and control groups. The intervention was an opinion-leader strategy based on patterns of social influence observed in empirical studies of innovation. We surveyed the obstetric clinicians--obstetricians, family practitioners, and nurse midwives--in each hospital to document their knowledge, attitudes, and beliefs, and identify their opinion leaders. Using triangulation, we also interviewed obstetric nurse managers in each hospital. During the study year, 12/96 through 11/97, we recruited the 9 opinion leaders, convened two workshops to assist development of their implementation strategies, provided written materials for clinicians and mothers, and monitored activities at the sites. Using multiple regression, the analysis compared the difference in breast-feeding rates between the prior baseline year and the study year in intervention vs. control hospitals. Besides the intervention group, matched hospital group, and study year, the analysis included independent variables to control for characteristics of mothers and clinicians likely to influence breast-feeding rates. A GEE (generalized estimating equation) procedure adjusted for clustering of mothers within clinician and clinicians within hospitals.
Results: Clinicians responding to the survey (59%) had a high degree of knowledge about the value of breast-feeding and management of associated problems. At each hospital, the 3 obstetric specialities agreed on their identification of the opinion leader, and the nurse manager agreed with the clinicians. All the opinion leaders also held administrative positions. The opinion leaders developed and implemented activities that included department meetings, working with nursing staff and pediatricians, changes in hospital policies, and earlier discussions with mothers. In the multivariate analysis, breast-feeding rates in the intervention hospitals, however, did not differ significantly from those in the control hospitals. This pattern pertained to mothers' intentions to breast feed exclusively and to their intentions to combine breast and formula feeding. Aside from the variables related to the intervention's effectiveness, the independent variables on mothers' and clinicians' characteristics were significantly associated with mothers' breast-feeding intentions.
Conclusions: The project demonstrated the feasibility of identifying common opinion leaders among multidisciplinary health professionals. Relying on nurse managers to identify opinion leaders and the overlap between opinion leaders and administrators merit further exploration. Though not significant, the results help to improve understanding of the circumstances in which different behavior-change strategies are effective. The opinion-leader strategy may be more effective in changing professional behavior than in changing organizational policies, community attitudes, or consumer behavior.