Effect of Accreditation in the Baby-Friendly Hospital Initiative on breastfeeding outcomes: A propensity score weighting approach

Abstract

<p>Background: There are numerous studies on the BFHI in the published literature; however, most studies have adopted a quasi-experimental study design owing to the nature of the treatment. There are a limited number of randomized trials and cluster-randomized controlled trial on the BFHI. Current literature review suggest that evidence on the effectiveness of the BFHI remains inconclusive. Furthermore, no studies have examined the effect of partial implementation of the BFHI in comparison to fully accredited BFHI.</p><p>Objectives/Hypothesis: To estimate treatment effects of the Baby-Friendly Hospital Initiative by comparing two cohorts before and after implementation of the BFHI using propensity score weighting and to examine differences in the effect on breastfeeding between hospitals that have begun the certification process and hospitals with BFHI designation.</p><p>Methods: A repeated prospective cohort study design was used. Propensity score weighting of baseline data was used to estimate the effects of the BFHI on breastfeeding outcomes. Participants were recruited from postnatal wards, 2369 mother–infant pairs were included from two cohorts, before and after implementation of the BFHI. The cohort after BFHI implementation was recruited from four study sites, with two designated BFHs and two in the process of accreditation that had received a Certificate of Commencement.</p><p>Results: Propensity score weighting achieved significant improvements in the balance of covariates of the two cohorts. In the adjusted model, full accreditation of the BFHI was associated with a 42% (95% confidence interval [CI]: 0.34–0.52) lower risk of weaning from any breastfeeding; partial accreditation was associated with a 39% (95% CI: 0.32–0.48) lower risk of weaning from any breastfeeding. When comparing accreditation stages of hospitals in the risk of weaning from exclusive breastfeeding, the adjusted model showed a 149% (95% CI: 1.01–2.21) lower risk of weaning from exclusive breastfeeding (p=0.04) in fully accredited BFHI hospitals compared with partial BFHI implementation.</p><p>Conclusions: Maternity institutions are encouraged to implement the BFHI and become fully accredited BFHI hospitals to improve breastfeeding outcomes. We observed potential benefits even with partial BFHI implementation.</p>