More than 30 years ago the book Guide to Effective Care in Pregnancy and Childbirth, systematically analysed the evidence for maternity care practices. In this book, Enkin et al. (1989) categorised interventions as: beneficial, ineffective, harmful, or unknown. In the last few decades, some routine practices have been abandoned (e.g., enemas and shaving in labour). However, resources have predominantly focussed on finding and implementing evidence-based interventions. While little attention has been paid to formal de-implementation (DI) of maternity interventions that are ineffective, harmful, wasteful or lack evidence at all.
A quick search of the literature finds zero research articles on DI in maternity care. However, there are examples of DI occurring. In 2020, the UK Saving Babies Lives Bundle required a second clinician to assess the fetal heart rate (using intermittent auscultation) each hour during labour. Citing lack of evidence, among other concerns, Feeley (2020) led a successful action to have this requirement removed from the bundle. You can read more about it here: https://www.all4maternity.com/de-implementing-fresh-ears/
In a previous post, we have argued that central fetal monitoring is a technology ripe for removal. See more here: Central fetal monitoring – time to de-implement? – Transforming Maternity Care Collaborative But identifying what needs to be removed is easier than the complex process of abandoning practices that are deeply embedded.
Which interventions should be de-implemented?
Norton & Chambers (2020) offer a useful analysis of the barriers to DI of healthcare interventions – and provide recommendations for success. First, they suggest interventions be classified as ineffective, contradicted, mixed or untested – to help prioritise those for DI (Norton & Chambers, 2020). Ineffective interventions have high-level evidence that demonstrates lack of benefit, or more harm than benefit. Contraindicated interventions are those for which more recent higher quality research contraindicates previous lower-level research that reported benefit. Mixed interventions have similar levels and amounts of evidence both for and against the effectiveness of the intervention. Untested interventions mean there is little or no evidence at all because of lack of research. While simple interventions (e.g., tests) may be easy to remove, complex interventions that have been used for a long period of time are more challenging (Norton & Chambers, 2020).
Barriers and facilitators
Factors that affect DI success include how healthcare users respond. For example, pregnant women may be anxious about not having access to an intervention they previously had or believe they need (Norton & Chambers, 2020). Additionally, social and cultural beliefs (i.e., that high-technology birth is safer than low-technology birth) may affect peoples’ willingness to accept alternatives (Norton & Chambers, 2020). Therefore, maternity consumer response needs to be planned for and managed as part of a co-designed DI strategy.
Health professionals may experience anxiety about not delivering an intervention, either based on a previous negative outcome, or fear of medical malpractice litigation (Norton & Chambers, 2020). This means the DI strategy should identify, mitigate or alleviate health professionals’ concerns. Organisations may resist DI due to liability concerns, loss of revenue or competitive advantage (Norton & Chambers, 2020). At a systems-level, application of financial disincentives for organisations that use of inappropriate interventions may be needed (Norton & Chambers, 2020). The key challenge is working out how to stop habitual behaviour that has become embedded because of training, usage, convenience or alignment with organisational policies or funding incentives (Davidson, 2017).
Ways to de-implement healthcare interventions
Norton & Chamber (2020) clarify that not all DI is removing an intervention. Instead it may mean replacing, reducing, or restricting its use. Replacing an intervention means stopping the inappropriate intervention and replacing it with an appropriate evidence-based intervention. For example, replacing continuous fetal heart rate monitoring with intermittent auscultation (Al Watter et al., 2021), or directed pushing in second stage with undirected pushing (Lemos et al., 2017). Reducing an intervention means changing the frequency or intensity of an intervention. For example, reducing the proportion of women who have a repeat caesarean section after one prior caesarean section. Restricting an intervention narrows its use to a target group, health professional or healthcare setting. For example, restricting the use of induction of labour to a narrow list of evidence-based clinical indications.
De-implementation in maternity care
To transform maternity care for the benefit of mothers and babies, it is time to shift our attention to de-implementation. In their recent call to action, Kennedy et al., (2018) stress that resources are being wasted on continued research and health service investment in ineffective or harmful practices – rather than supporting improved access to, and sustainability of, evidence-based, high quality care. However, practices that are deeply ingrained in an organisation’s infrastructure may be extremely difficult to de-implement (Montini & Graham, 2015) even when the evidence is persuasive (Upvall et al., 2019; Van Bod et al., 2017).
But it is possible.
Although not termed ‘de-implementation’, the World Health Organisation & UNICEF’s (2018) Baby Friendly Hospital Initiative demonstrates how inappropriate practices (e.g., routine separation, timed breastfeeds, liberal use of artificial formula) can be successfully replaced / restricted by use of evidence-based practices (e.g., skin-to-skin, rooming-in, unrestricted breastfeeding, and artificial formula for a medical indication only).
Norton, W.E., & Chambers, D.A. (2020). Unpacking the complexities of de-implementing inappropriate health interventions. Implementation Science, 15, 2. https://doi.org/10.1186/s13012-019-0960-9
Al Wattar, B.H. et al. (2021). Effectiveness of intrapartum fetal surveillance to improve maternal and neonatal outcomes: a systematic review and network meta-analysis. CMAJ, 193(14), e468-e477. https://doi.org/10.1503/cmaj.202538
Davidson, K. W., Ye, S., & Mensah, G. A. (2017). Commentary: de-implementation science: a virtuous cycle of ceasing and desisting low-value care before implementing new high value care. Ethnicity & Disease, 27(4), 5.
Enkin et al. (1989). Guide to Effective Care in Pregnancy and Childbirth.
Feeley, C. (2020). De-implementing ‘fresh ears’. Retrieved from: https://www.all4maternity.com/de-implementing-fresh-ears/
Kennedy, H. P., Cheyney, M., Dahlen, H. G., Downe, S., Foureur, M. J., Homer, C. S. E., . . . Renfrew, M. J. (2018). Asking different questions: A call to action for research to improve the quality of care for every woman, every child. Birth, 45(3), 222-231. https://doi.org/10.1111/birt.12361
Lemos et al. (2017). Pushing/bearing down methods for the second stage of labour. Cochrane Database of Systematic Reviews. Retrieved from: https://www.cochrane.org/CD009124/PREG_pushing-methods-second-stage-labour
Montini, T. & Graham, I.D. (2015). “Entrenched practices and other biases”: unpacking the historical, economic, professional, and social resistance to de-implementation.” Implementation Science, 10(1), 24.
Upvall, M. J., A. M. Bourgault, C. Pigon and C. A. Swartzman (2019). Exemplars Illustrating De-implementation of Tradition-Based Practices. Critical Care Nurse, 39(6), 64-69.
van Bodegom-Vos, L., F. Davidoff and P. J. Marang-van de Mheen (2017). Implementation and de-implementation: two sides of the same coin? BMJ Quality & Safety, 26(6), 495.
World Health Organization & UNICEF (2018). Implementation Guidance: Protecting, Promoting and Supporting Breastfeeding in Facilities Providing Maternity and Newborn Services: The Revised Baby-Friendly Hospital Initiative. WHO.