Complex interventions, healthcare innovations, and the dynamics of implementation An introduction to Normalization Process Theory Carl May

My wonderful collaborator and co-investigator in building NPT….

Grants RES 000-27-0084; RES 189-25-0003; RES 062-23-3274 Acknowledgements: Frances Mair, Jane Gunn, Mark Johnson, Susi Lund, Anne MacFarlane, Victor Montori, Catherine O’Donnell, Elizabeth Murray, Catherine Pope, Tim Rapley, Anne Rogers, Alison Richardson, Nilay Shah, Shaun Treweek Grants RES 000-27-0084; RES 189-25-0003; RES 062-23-3274

There is nothing so practical as a good theory Kurt Lewin NIHR There is nothing so practical as a good theory Kurt Lewin

Implementation is a problem

What is implementation? NIHR What is implementation? Implementation includes any deliberately initiated attempt to introduce new, or modify existing, patterns of collective action in health care or some other formal organizational setting. Deliberate initiation means that an intervention is: institutionally sanctioned; formally defined; consciously planned; and intended to lead to a changed outcome. Participants may seek to modify the ways that people think, act and organize themselves or others, they may seek to initiate a process with the intention of creating a new outcome.

What is implemented? Interventions NIHR What is implemented? Interventions may be intended to change behaviour and its intended outcomes (e.g. strategies for making ‘expert patients’; or using telemedicine systems) may be intended to change expertise and actions (e.g. devices; or decision-making tools and clinical guidelines) may be intended to change the procedures enacted to achieve goals. (e.g. electronic health records, ordering systems)

NIHR More than 60 theories, models, and frameworks relevant to implementation are available to practitioners and researchers (Tabak et al, 2012) Heavy emphasis on attributes of organizations and policy environments (inner and outer contexts), reflects influence of diffusion models. Heavy emphasis on individual differences (attitudes and intentions), reflects influence of psychological individualism.

Why build new theory? robust social science theories already explain individual differences in attitudes to new technologies and practices (e.g. Theory of Planned Behavior) the flow of innovations through social networks (e.g. Diffusion of Innovations Theory and related frameworks). reciprocal relations between people and things (e.g. Actor Network Theory) There seemed to be a peculiar absence of a theory of implementation, embedding and integration: an implementation theory shaped hole in implementation science.

Human behaviour is often interesting and exotic

Normalisation Process Theory Provides a conceptual toolkit to understand, explain and organise ‘implementation processes’ as a technical problem of practice. Provides a way to understand ‘implementation processes’ as central elements of societies organised around ‘projects’ made up of temporary assemblages of agents and ensembles of social practices.

NPT Formal Theory grounded in >100 empirical studies

How users interact with interventions – characterizes capability Interactional workability: defines how a complex intervention is practically operationalized by the people using it Skill-set workability: defines the distribution and conduct of work associated with a complex intervention in a division of labour Relational integration: defines knowledge and work about a complex intervention is mediated and understood within networks. Contextual integration: the realization of resources of a complex intervention within an organizational domain.

Supporting implementation design Hoberg, A. et al., Feasibility evaluation of Interpersonal and Social Rhythm Group Therapy Delivery Model Archives of Psychiatry In Press

Core constructs: characterize how social mechanisms focus agentic investments – characterizes agentic contribution Coherence: defines and organizes the components of a complex intervention Collective Action: defines and organizes the enacting of a complex intervention Cognitive Participation: defines and organizes the people implicated in a complex intervention Reflexive Monitoring: defines and organizes assessment of the outcomes of a complex intervention

Quality improvement collaborative for depression (13 primary care MDTs, Netherlands) Coherence: The stepped-care model offered clinicians a technique for shared understanding on depression (who is severely and non severely depressed). Cognitive participation: The new low intensity stepped-care treatment options fitted well into the primary care perspective. Collective action: The possibility to tailor the stepped-care model to the local setting, and to train staff to apply the stepped-care interventions was important, but poor organizational infrastructures and lack of funding of the new low intensive interventions. Reflexive monitoring: Improved motivation because outcome measurement can structure and advance care for individual patients. But absence of supportive systems (ICT, reminder systems) or staff. Franx G, et al,. Implementing a stepped care approach in primary care Implement Sci 2012, 7(8)

It’s all about the work What is the work? (How is a practice made coherent by its users?) Who does the work? (How do people and groups come to participate into a complex intervention?) How does the work get done? (How is a complex intervention enacted in practice?) Why did the work happen like that? (How is a complex intervention monitored by its users?)

Relationships between capability, contribution and context* Capacity: social structural resources (norms, roles) available to agents Contribution: agency expressed through coherence; participation; action; monitoring Potential: social cognitive resources (readiness, commitment) available to agents Capability: workability and integration of the implementation object

Normalisation Process Theory: core components NIHR Capacity and Potential - the social and cognitive resources on which agents draw (May 2013a,b) 1. The incorporation of a complex intervention within a social system depends on structural effects on agents’ capacity to co-operate and co-ordinate their actions. 2. The translation of potential into collective action depends on agents’ potential to enact the complex intervention. Contribution - the mechanisms through which agency is expressed (May & Finch 2009) 3. The implementation of a complex intervention depends on agents’ contributions that carry forward in time and space. 3.1 Implementation of a complex intervention is dependent on work that defines and organizes a practice as a cognitive and behavioral ensemble. 3.2 Implementation of a complex intervention is dependent on work that defines and organizes the actors implicated in a practice. 3.3 Implementation of a complex intervention is dependent on work that defines and operationalizes a practice. 3.4 Implementation of a complex intervention is dependent on work that defines and organizes the everyday understanding of a practice. Capability - attributes of the relations between people and things (May 2006, May et al 2007a,b) 4. The capability of agents to employ a complex intervention depends on its workability and integration within a social system. 4.1 Workability of a complex intervention depends on the extent to which it confers an interactional advantage to the user. 4.2 Integration of a complex intervention depends on the extent of users’ confidence in it. 4.3 Workability of a complex intervention depends on the extent to which it is calibrated to a skill-set in a division of labor. 4.4 ntegration of a complex intervention depends on the extent to which it confers an advantage in executing tasks. Normalisation Process Theory: core components

(Social-structural resources available to agents) Capability Contribution (What agents do to implement a complex intervention) Capacity (Social-structural resources available to agents) Capability (Possibilities presented by the complex intervention) Potential (Social-cognitive resources available to agents)

3. Applying npt in practice

Susi Lund, Alison Richardson & Carl May Advance Care Plans at End of Life: Systematic review of and development of a conceptual model of implementation processes informed by NPT

An ACP is an interactional process between a patient, significant others, and clinicians. Cognitive and communicative incapacity, and loss of self at end of life, are acknowledged as a future risk. Patient’s preferences about clinical and other actions are expressed, negotiated, acknowledged and recorded by other participants in the expectation that they will be acted upon if necessary.

PLoS ONE 10(2): e0116629. doi:10.1371/journal.pone.0116629 Fig 1. PRISMA flowchart. Lund S, Richardson A, May C (2015) Barriers to Advance Care Planning at the End of Life: An Explanatory Systematic Review of Implementation Studies. PLoS ONE 10(2): e0116629. doi:10.1371/journal.pone.0116629 http://127.0.0.1:8081/plosone/article?id=info:doi/10.1371/journal.pone.0116629

We consider ACPs as an implementation problem, and ask: what factors promote or inhibit the routine incorporation of ACPs in clinical practice? Systematic literature review: focuses on existing research in different healthcare systems.

Fig 1. PRISMA flowchart. Lund S, Richardson A, May C (2015) Barriers to Advance Care Planning at the End of Life: An Explanatory Systematic Review of Implementation Studies. PLoS ONE 10(2): e0116629. doi:10.1371/journal.pone.0116629 http://127.0.0.1:8081/plosone/article?id=info:doi/10.1371/journal.pone.0116629

Table 2. Coding frame and Taxonomy Items. Lund S, Richardson A, May C (2015) Barriers to Advance Care Planning at the End of Life: An Explanatory Systematic Review of Implementation Studies. PLoS ONE 10(2): e0116629. doi:10.1371/journal.pone.0116629 http://127.0.0.1:8081/plosone/article?id=info:doi/10.1371/journal.pone.0116629

We undertook an explanatory review of literature describing the operationalization of ACPs The principal inclusion criterion was that papers should report the implementation of interventions intended to support Advance Care Planning in adult healthcare settings. We excluded interventions aimed at children and those with mental health problems.

Operational contexts are under pressure (other work competes). Patient trajectories are uncertain (prognostication is difficult). Negotiations have unpredictable outcomes. (emotional complexity is stressful) Advanced Care Plans may not be actioned (patients and staff may be overtaken by events).

Fig 2. Operationalizing Advance Care Plans: Facilitators. Lund S, Richardson A, May C (2015) Barriers to Advance Care Planning at the End of Life: An Explanatory Systematic Review of Implementation Studies. PLoS ONE 10(2): e0116629. doi:10.1371/journal.pone.0116629 http://127.0.0.1:8081/plosone/article?id=info:doi/10.1371/journal.pone.0116629

Resources: slideshares; translational framework; technology adoption readiness scale; on-line toolkit

www.normalizationprocess.org Resources: key papers May C, Finch T. Implementation, embedding, and integration: an outline of Normalization Process Theory. Sociology 2009; 43:535-54. Available here May C, et al. Development of a theory of implementation and integration: Normalization Process Theory. Implementation Science 2009; 4. Available here. May C, Towards a general theory of implementation. Implementation Science 2013, 8:18 Available here www.normalizationprocess.org

Thank you! Carl May | Faculty of Health Sciences |University of Southampton | Southampton SO17 1BJ, UK | Tel ++ 44 (0)23 8059 7957 | Skype CarlRMay | Twitter @CarlRMay | email c.r.may@soton.ac.uk