Gill Thomson,1 Heather Trickey2
1. Senior Research Fellow, MAINN, University of Central Lancashire, Preston, UK
2. Centre for the Development and Evaluation of Complex Interventions for Public Health Improvement (DECIPHer), Cardiff School of Social Sciences, Cardiff University, UK; Research Manager, NCT, Alexandra House, Oldham Terrace, London, UK
Disclosure: No potential conflicts of interest.
Received: 07.10.13 Accepted: 10.12.13
Citation: EMJ Gyn Obs. 2013;1:15-22.
Policymakers from developed countries who are looking to commission breastfeeding peer support (BPS) services have every cause to be puzzled as to whether or not they can improve continuation rates. On the one hand, BPS interventions are internationally recognised as having the potential to contribute to improving breastfeeding durations.1 A recent Cochrane review found that additional support from lay and professional supporters can have an impact on rates,2 and UK-based qualitative studies suggest that BPS can encourage and enable women to breastfeed for longer periods.3,4 In the UK, peer support for breastfeeding forms part of NHS commissioning guidance.5 On the other hand, a recent meta-regression of BPS randomised controlled trials (RCTs) found little evidence that BPS interventions improve breastfeeding durations in high-income countries6 and concluded that peer support for breastfeeding was ‘unlikely to be effective’ in the UK. This paper highlights issues of intervention design and implementation that problematise interpretation of trial data drawn from the meta-regression analysis within high income countries. The paper then goes on to consider the potential for alternative approaches to review evidence for BPS, highlighting the need to integrate insights from qualitative research studies. Drawing on findings of a preliminary scoping review, we make the case for a shift towards a realist interpretation of the evidence base. We argue that a realist approach would allow findings emergent from different methodological traditions to be meaningfully integrated and the theoretical basis for BPS to be explored and tested through the construction of context-mechanism-outcome configurations. We believe this will provide a firmer basis for future intervention design and for the development of theoretically-driven evaluation studies, leading to improved clarity for delivery organisations and commissioning agencies. We contend that policy makers and researchers need to stop merely asking ‘does BPS work?’ and look towards approaches which enlighten ‘what works for whom, in what circumstances, in what respects, and how?’
Download (PDF, Unknown)