Scaling up an intervention to protect preterm infants from neurodevelopmental disabilities — findings from a qualitative process evaluation comparing standard with enhanced quality improvement support packages for maternity units in England

BACKGROUND: A quality improvement strategy (PReCePT) was used in a standard and enhanced format to scale up a clinical intervention (administering magnesium sulphate to women in preterm labour) across all maternity units in England to protect prematurely born infants from neurodevelopmental disabilities. Formal evaluations reported the effectiveness of the standard package alone in increasing the administration of magnesium sulphate. In this paper, we focus on the findings of the process evaluations, using normalisation process theory to explain how different implementation contexts generated the observed outcomes relating to normative and relational restructuring and sustainment. METHODS: Interviews were conducted with key individuals in implementation of leadership positions nationally and locally. Interviews were analysed initially using the framework method. We then engaged recursively with NPT constructs to generate generalisable insights with pragmatic applicability in other settings. RESULTS: In total, 72 interviews were conducted with good representation from units across England and staff from the National Academic Health Science Network. We found that all units irrespective of whether they received a standard or enhanced QI package were successful in the ‘normative restructuring’ of their setting to enable magnesium sulphate to be administered. This suggests that this implementation outcome is necessary to achieve improvements. However, it may not be sufficient to sustain the changes once additional resources have been withdrawn. Sustainment, our findings suggest, required ‘relational restructuring’ to accommodate altered workflows and facilitate the sharing of responsibilities and tasks in daily practice. Relational restructuring was more likely to have been achieved units receiving enhanced QI support but also happened in units with standard QI support, especially in those where perinatal team working was already well established. CONCLUSION: Unlike other large QI-focused spread-and-scale programmes which failed to show any impact on outcomes, the PReCePT programme in both the enhanced and standard support packages led to improvements in the uptake of magnesium sulphate. The findings suggest that QI programmes interact with the enabling factors, such as strong interprofessional team working, already present in the setting. A standard package with minimal support was therefore sufficient in settings with enabling factors, but enhanced support was required in units where these were absent.