Implementation of a complex intervention to improve care for patients whose situations are clinically uncertain in hospital settings: a multi-method study using normalisation process theory

Halle Johnson, Emel Yorganci, Catherine J. Evans, Stephen Barclay, Fliss E. M. Murtagh, Deokhee Yi, Wei Gao, Elizabeth L. Sampson, Joanne Droney, Morag Farquhar, Jonathan Koffman

Research output: Contribution to journalArticlepeer-review

1 Citation (Scopus)
28 Downloads (Pure)


Purpose: To examine the use of Normalisation Process Theory (NPT) to establish if, and in what ways, the AMBER care bundle can be successfully normalised into acute hospital practice, and to identify necessary modifications to optimise its implementation.

Method: Multi-method process evaluation embedded within a mixed-method feasibility cluster randomised controlled trial in two district general hospitals in England. Data were collected using (i) focus groups with health professionals (HPs), (ii) semi-structured interviews with patients and/or carers, (iii) non-participant observations of multi-disciplinary team meetings and (iv) patient clinical note review. Thematic analysis and descriptive statistics, with interpretation guided by NPT components (coherence; cognitive participation; collective action; reflexive monitoring). Data triangulated across sources.

Results: Two focus groups (26 HPs), nine non-participant observations, 12 interviews (two patients, 10 relatives), 29 clinical note reviews were conducted. While coherence was evident, with HPs recognising the value of the AMBER care bundle, cognitive participation and collective action presented challenges. Specifically: (1) HPs were unable and unwilling to operationalise the concept of ‘risk of dying’ intervention eligibility criteria (2) integration relied on a ‘champion’ to drive participation and ensure sustainability; and (3) differing skills and confidence led to variable engagement with difficult conversations with patients and families about, for example, nearness to end of life. Opportunities for reflexive monitoring were not routinely embedded within the intervention. Reflections on the use of the AMBER care bundle from HPs and patients and families, including recommended modifications became evident through this NPT-driven analysis.

Conclusion: To be successfully normalised, new clinical practices, such as the AMBER care bundle, must be studied within the wider context in which they operate. NPT can be used to the aid identification of practical strategies to assist in normalisation of complex interventions where the focus of care is on clinical uncertainty in acute hospital settings.
Original languageEnglish
Article numbere0239181
JournalPLoS One
Issue number9
Publication statusPublished - 16 Sep 2020

Cite this