Assessing the effectiveness of an algorithm decision tool in decreasing non-medically indicated supplementation among newborns in a hospital setting

STUDY AIM: This retrospective quantitative study assessed the effectiveness of an algorithm decision tool and hospital staff training clinics, designed to decrease supplementation of newborns at Alameda County Medical Center (ACMC), by comparing the in-hospital supplementation prevalence rates in the periods before and after the decision tool and training sessions were implemented. BACKGROUND: Supplementation, bottle feedings given in place of or in addition to breastfeeding, of newborns in the hospital setting may negatively impact the duration of breastfeeding. ACMC staff designed both an algorithm decision tool and training clinics for nurses and providers to decrease in-hospital supplementation of newborns that is non-medically indicated. This intervention was part of their larger goal to be implementing current best evidence in making decisions for the health care of maternal and infant patients, to achieve an in-hospital supplementation prevalence rate close to 22%, and to achieve a Baby Friendly Hospital accreditation granted by the World Health Organization. METHODS: Eligible pools were established of women who gave live birth at ACMC during the 11-months preceding or the 10-months following the decision tool and training implementation, and who met exclusion criteria. From these eligible pools of 1063 and 948 respectively, sample populations of 115 each were randomly selected. Data was collected from notations in the medical record of birth mothers to determine whether supplementation or exclusive breastfeeding occurred in-hospital, and why supplementation occurred if this could be ascertained. Further exclusions were found in the records. Primary language spoken by the mother was also noted. Collected data from each birth mother's medical record was cross-referenced with the corresponding infant medical record to ensure reliability of notations. RESULTS: A significant reduction (p=0.04) in the prevalence rate of in-hospital non-medically indicated supplementation was found between the pre-tool/training prevalence rate of 53% (n=100) and the post-tool/training rate of 38.6% (n=101). Prevalence rates for non-English speaking women were also calculated and no significant difference was found between this sub-population and the sample population either pre-tool/training (p=0.4) or post-tool/training (p=0.9). CONCLUSIONS: The algorithm decision tool and training was significantly effective in decreasing non-medically indicated in-hospital supplementation of newborns at ACMC. This significant reduction provides insight to ACMC staff as they assess progress in practicing evidence-based medicine in the maternal and child health department and continue their intervention of training staff in using the algorithm decision tool that is required to achieve their goals.