years and future directions Michel Wensing, Paul Wilson and Anne - - PowerPoint PPT Presentation

https://implementationscience.biomedcentral.com/

Implementation Science: the first 10 years and future directions

Michel Wensing, Paul Wilson and Anne Sales

https://implementationscience.biomedcentral.com/

Eccles and Mittman. Implement Sci 2006; 1:1.

https://implementationscience.biomedcentral.com/

Journal mission Implementation Science is an open access, peer-reviewed online journal that aims to publish research relevant to promoting the uptake of research findings into healthcare practice and health policy.

Foy et al. Implement Sci 2015;10:51.

https://implementationscience.biomedcentral.com/

Key facts

• Founded in 2006 by Martin Eccles and Brian Mittmann
• Published by Biomedcentral, now part of Springer Nature, as open access

journal

• Not formally linked to any particular organisation (material support from

VA until 2014)

• Current editors-in-chief: Anne Sales, Michel Wensing, Paul Wilson
• Group of Editors and international Editorial Board
• Currently > 700 submissions per year, < 25% acceptance rate

https://implementationscience.biomedcentral.com/

Manuscripts submitted / published

11 100 59 87 175 252 254 441 529 644 764 100 200 300 400 500 600 700 800 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Submitted Published

https://implementationscience.biomedcentral.com/

Publications and citations (23nov2016)

• Total publications: 1213
• Total citations: 16480
• Average citations per

article: 13.59

Publications per year Citations per year

https://implementationscience.biomedcentral.com/

Impact factor

• 2016 IF = 3.201 (5 year IF = 4.325)
• Ranked 15/87 in Health Care Sciences & Services

(SCI)

• Ranked 5/74 Health Policy & Services (SSCI)

https://implementationscience.biomedcentral.com/

Editorials

• De-implementation (Prasad & Ionnanidis)
• Economics of implementation (Severens )

https://implementationscience.biomedcentral.com/

Program (14.30-15.45)

• Paul Wilson: Evolving scope and methods
• Anne Sales: Theories and frameworks
• General discussion

https://implementationscience.biomedcentral.com/

10 years of evolving scope and methods

Paul Wilson Deputy Editor in Chief, Implementation Science Alliance Manchester Business School paul.wilson@manchester.ac.uk

https://implementationscience.biomedcentral.com/

Sustained growth over last ten years

100 200 300 400 500 600 700 800 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Submitted Published

https://implementationscience.biomedcentral.com/

Our current scope

• Research examining the implementation of evidence-based

practices or policies or the de-implementation of those demonstrated to be relatively ineffective or even harmful

• Study of processes, fidelity and context
• Study of influences on provider, patient and organisational

behaviour

• Includes population health interventions

Foy et al. Implement Sci 2015;10:51.

https://implementationscience.biomedcentral.com/

So not….

• Research establishing the effectiveness of novel clinical,
• rganisational, public health or policy interventions
• Quality improvement or service evaluation
• Weak designs
• Doing implementation or lessons learned

https://implementationscience.biomedcentral.com/

General guiding principles

• Address a clear question using an appropriate design
• Build on existing knowledge
• Increasingly priority given to substantial, rather than marginal,

contributions to the field

https://implementationscience.biomedcentral.com/

Implementation effectiveness

• Randomised controlled trials

– Cluster RCTs

• Non-randomised controlled trials (NRCTs)
• Controlled before-after studies
• Interrupted time series (difference in difference, repeated

measures)

https://implementationscience.biomedcentral.com/

Cluster RCTs in Implementation Science

5 10 15 20 25 30 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 Findings Protocols

https://implementationscience.biomedcentral.com/

Hybrid design types

I. Effects of a clinical intervention on relevant outcomes while

• bserving and gathering information on implementation
• II. Dual testing of clinical interventions and implementation

strategies

• III. Testing an implementation strategy while observing and

gathering information on the clinical intervention's impact on relevant outcomes

Curran et al. Med Care 2012;50:217-26.

https://implementationscience.biomedcentral.com/

Prospective trials registration

• Now only normally consider for publication trials that have

been registered prior to enrolment of the first participant/ cluster

• Continue consider retrospectively registered trials on a case

by case basis

• Should we encourage prospective registration of other study

designs?

https://implementationscience.biomedcentral.com/

You damned positivists

But not just about trials

https://implementationscience.biomedcentral.com/

Study of processes, fidelity and context

Moore et al. BMJ 2015; 350:h1258.

https://implementationscience.biomedcentral.com/

https://implementationscience.biomedcentral.com/

https://implementationscience.biomedcentral.com/

Enhancing reporting

• Variation in the standards of reporting of the research that we

publish

– CONSORT – PRISMA

• Quality of intervention description remains an issue

– TIDier – WIDER – StaRI

https://implementationscience.biomedcentral.com/

Protocol publication

• Fit our journal scope and inclusion criteria for rigorous study

designs AND

• Have received funding from a nationally or internationally

recognised research agency AND

• Have received appropriate ethics review board approval AND
• Submitted within three possible time points:

– 1) WITHIN 3 months of ethics approval – 2) PRIOR to enrolment of the first participant/ cluster – 3) BEFORE the end of participant/ cluster recruitment

https://implementationscience.biomedcentral.com/

In summary

• 10 years of sustained growth
• Field and scope has evolved and as submissions have risen,

we have become more selective in what we publish.

• Scientific quality and transparency key drivers in selection
• Priority given to research that offers significant contributions

to knowledge, methodology or thinking

https://implementationscience.biomedcentral.com/

10 years of theories and frameworks in Implementation Science

Anne Sales Editor in Chief, Implementation Science

https://implementationscience.biomedcentral.com/

Theory vs. framework

• Theories contain constructs

– Sub-particles invoking specific phenomena

• Posit relationships among

constructs

– X affects Y in some specific way

• Graphic representation

generally includes arrows which describe relationship

• Frameworks contain

constructs

• Constructs are categorized
• No formal relationships

among constructs are proposed

• Graphic representation (if

any) is mostly heuristic

• Checklists

https://implementationscience.biomedcentral.com/

Searching on “framework” in IS

• 963 results

– Most are papers or conference proceedings that explicitly explore a specific framework – Some are new frameworks proposed for use in research

• Many of these are very purpose-built and focus on specific topical areas

– Most use frameworks but few contribute new insights – Some are original new frameworks

• Some of these are consolidations of existing frameworks and/or theories
• Notable examples

– Damschroder et al. Implementation Science 2009 4:50 » Consolidated Framework for Implementation Research framework – Flottorp et al. Implementation Science 2013 8:35 » Tailored Implementation for Chronic Disease checklist

https://implementationscience.biomedcentral.com/

Searching on “theory” in IS

• 785 results

– Many are papers or conference proceedings that describe using or testing theory rather than building or creating theory

• Many invoke the term theory as a building block for designing

interventions or instruments

• Many of these use frameworks rather than theories for this kind of design
• Primary approach is to use a framework or theory to assess barriers

and/or facilitators (primarily barriers) then design or tailor an intervention to address these

• Often described as “theory-based intervention design”

– Some are articulations of new or extended theories

• May Implementation Science 2013 8:18
• May et al. Implementation Science 2009 4:29

https://implementationscience.biomedcentral.com/

Lots of results

• Suggest that use of a framework or theory is seen as

important in implementation research

– Not a uniform perception – Two of the first papers published in IS debate the utility of theory in implementation research

• The ICEBeRG group Implementation Science 2006 1:4
• Bhattacharyya et al. Implementation Science 2006 1:5

– Evidence for this position is not strong

• Frameworks are not well distinguished from theory

– Suggests some degree of confusion – Unclear whether there is any way to “test” a framework – Testing theory is not trivial but theoretically possible – The essential “checklist” nature of frameworks is not recognized

https://implementationscience.biomedcentral.com/

A brief digression: the Theoretical Domains Framework

• Multi-year effort by consortium of UK-based health

psychologists, implementation researchers and health services researchers

• Deconstructs 33 health psychology theories into 12-14

domains

– 93 constructs – 14 domains in current iteration

• Original papers published elsewhere
• Series on TDF published in IS 2012

– http://www.biomedcentral.com/collections/tdf

https://implementationscience.biomedcentral.com/

Three important recent papers

• Nilsen Implementation Science 2015 10:53

– Provides a taxonomy of theories and frameworks

• Chaudoir et al. Implementation Science 2013 8:22

– Systematic review of factors affecting implementation of health innovations; provides an overview of constructs with reference to theory or framework and a review of instruments to measure constructs when available

• Tabak et al. American Journal of Preventive Medicine 2012

http://dx.doi.org/10.1016/j.amepre.2012.05.024

– Systematic review of 61 frameworks used in dissemination and implementation research; provides classification by level (individual,

• rganizational, societal) and by dissemination vs. implementation

https://implementationscience.biomedcentral.com/

Nilsen’s taxonomy

• Five categories for classifying theories, models and frameworks

– Process models

• How implementation gets done– typically in steps or stages, describing and/or guiding

the process of implementation

– Determinant frameworks

• Specify constructs or factors that affect implementation process/success– usually

classified within categories or domains; goal is to understand and/or explain implementation outcomes

– Classic theories

• Theories from the disciplines that underlie the synthetic field of implementation

research– many constructs derive from these theories, which are not unique to implementation research

– Implementation theories

• Theories developed by implementation researchers that posit relationships that guide

implementation practice and research

– Evaluation frameworks

• Frameworks that specify aspects of implementation that can be measured and should be

useful in evaluating whether an implementation effort is a success or not

https://implementationscience.biomedcentral.com/

Helpful guidance for moving forward in using frameworks, models and theories

• Multiple uses for different types of frameworks, models and

theories

• Quite likely that a theoretically based implementation project might

use at least 3 different frameworks

– Process model to guide process – Determinants framework to elicit barriers and/or facilitators to support implementation intervention design and/or tailoring – Evaluation framework to support evaluation

• Some models are hybrids of different kinds of frameworks

– This can be useful or it can be simply confusing

• Probably only theories can be tested

– But frameworks and/or models could be developed into theory and tested

https://implementationscience.biomedcentral.com/

But we still need evidence

• The ICEBeRG debate from 2006 is still an open question

– Is theory (or use of theory-informed frameworks) effective in supporting implementation? – Is theory useful, important or essential? – What kind of theory, articulated in what ways, and to what ends? – Important paper published in BMJQS

• Davidoff et al. 2015

http://qualitysafety.bmj.com/content/early/2015/01/23/bmjqs-2014- 003627.full?g=w_qualitysafety_open_tab

• New types of models/frameworks/theories may be emerging

– Adaptation and/or fidelity theories and frameworks one example

https://implementationscience.biomedcentral.com/

Questions and comments