UCSF Research Admin Board Presentation on CTSI Global Health Program
Handley 2012
1. A Few Implementation Frameworks for
Health Interventions in Global Context
Margaret Handley, PhD MPH
Associate Professor
UCSF Department of Epidemiology and
Biostatistics and
SFGH Division of General Internal Medicine,
Center for Vulnerable Populations
April 25, 2012
2. A Starting Point
“Many evidence-based innovations fail to
produce results when transferred to
communities in the global south, largely
because their implementation is untested,
unsuitable or incomplete”
T. Madon et al .2007
3. Implementation is Challenging
“Strategies required to deliver
good care in low-income
settings should recognize that
this will need to be co-
produced through
engagement, often over
prolonged periods and as part
of a directive but adaptive,
participatory, information-rich,
and reflective process”
-English M et al, 2011
4. Definitions
• A theory is a set of testable propositions that help us to explain
and predict phenomena, such as health behaviors. A theory is a
tool that allows one toinform and strengthen practical
solutions to old and emerging problems in public health.
• Planning models or frameworks exist at a macroscopic level;
they serve as an organizing framework for an entire health
promotion effort aimed at fostering reduction in a given
disease. Unlike theories, planning models are not made up of a
set of testable propositions. Rather, planning models serve as a
blueprint for building and improving intervention programs.
Crosby and Noar 2011
5. Conceptually Organizing Implementation Planning
Deciding Where you Are Starting From Affects the Choice of
Framework(s)
WHAT IS TARGET, CONTEXT, PLATFORMS/TOOLS
AVAILABLE?
TARGET TARGET
To gain an understanding of Are you starting with placing
a problem so as to develop an evidence-based
and test an intervention? intervention of known
What setting? How many? efficacy into a new setting?
How many tools? What setting? How many?
How many tools?
Behavior change theory,
intervention design, logic Some of each? Implementation-focused
model of the problem, models in which outcomes
logic model of may also include more
change>>behavior measures of success of
change outcomes- implementation
focused implementation
models
6. Foundations
• What are implementation “theories”
(models, frameworks)?
• How and why are they useful?
7. Why Implementation “Theory” or
Frameworks?
Theories, models, frameworks provide a
systematic method:
… for identifying, understanding,
operationalizing & evaluating the black box
phenomenon =
“IMPLEMENTATION”
8. Selecting a Theory/Model
• Multiple theories/models often needed
– Impact theories/models
• Specify the relationship hypothesized about
how implementation activities will contribute
to desired behavior change and overcome
barriers
– Process theories/models
• The ‘how’ of implementation
(planning, organization and scheduling)
Adapted from: Grol RP, Bosch MC, Hulscher ME, Eccles MP, Wensing M. Planning and studying
improvement in patient care: the use of theoretical perspectives. Milbank Q. 2007;85(1):93-138.
9. Selecting a Theory/Model cont.
• Multiple theories/models often needed
--Theories/models that focus on specific settings
or contexts (e.g. health systems, community
activation)
-- Theories/models that incorporate multiple
layers or multiple components to select from
10. Examples: Model focusing on Intervention
Development and Implementation
Select theory of Identify potential Select interventions
planned behavior strategies for that fit with planned
change achieving change strategies
(based on theory)
Assess fit with
initial theory
Evaluate effectiveness Launch intervention Identify intervention
of intervention, using identified tools tools that fit both
strategies, tools and strategies strategy and theory
12. Selected Models for Discussion
1. PRECEDE-PROCEDE (mostly the PRECEDE)
Predisposing, Reinforcing, and Enabling Constructs in
Educational/ environmental Diagnosis and Evaluation
2. Consolidated Framework for Implementation
Research (CFIR)
3. Promoting Action on Research Implementation
in Heath Research (PARIHS)
D
13. Conceptually Organizing Implementation Planning
Deciding Where you Are Starting From Affects the Choice of
Framework(s)
WHAT IS TARGET, CONTEXT, PLATFORMS/TOOLS
AVAILABLE?
TARGET TARGET
To gain an understanding of Are you starting with placing
a problem so as to develop PRECEDE-
PRECEDE- an evidence-based
and test an intervention? PROCEDE
PROCEDE intervention of known
What setting? How many? efficacy into a new setting?
How many tools? What setting? How many?
How many tools?
Behavior change theory,
intervention design, logic Some of each? Implementation-focused
model of the problem, models in which outcomes
logic model of may also include more
change>>behavior measures of success of
change outcomes- implementation
focused implementation
models
14. Selected Models for Discussion
1. PRECEDE-PROCEDE- the most widely used
planning model for the development and evaluation of
health promotion and policy programs in the world. Socio-
ecological.
•Organizing framework, multi-component, time-
sequenced activities that serve as a “blueprint,”
systematically guiding program developers
•Predisposing=
•Reinforcing=
•Enabling=
(Green & Kreuter, 1991, 2005).
15. Selected Models for Discussion
2. CFIR- Consolidated Framework for Implementation-
“A comprehensive practical taxonomy of constructs
that have an established evidence base in the
literature to enable implementation researchers to
see further through the complex array of influences
on implementation by bringing together constructs
across many different scientific disciplines into a
single framework for pragmatic and scientific
application”
Damschroeder and Hagedorn, 2011
16. Conceptually Organizing Implementation Planning
Deciding Where you Are Starting From Affects the Choice of
Framework(s)
WHAT IS TARGET, CONTEXT, PLATFORMS/TOOLS
AVAILABLE?
TARGET TARGET
To gain an understanding of Are you starting with placing
a problem so as to develop an evidence-based
and test an intervention? intervention of known
What setting? How many? efficacy into a new setting?
How many tools? What setting? How many?
How many tools?
Behavior change theory,
intervention design, logic Some of each? Implementation-focused
model of the problem, models in which outcomes
logic model of may also include more
change>>behavior Multi-level
Multi-level
implementation measures of success of
change outcomes- implementation
frameworks that implementation
focused implementation frameworks that
models address complex
address complex
interventions
interventions
21. Consolidated Framework for
Implementation Research (CFIR)
Intervention Outer Setting Intervention
(adapted)
(unadapted)
Adaptable Periphery
Core Components
Adaptable Periphery
Inner Setting
Core Components
Individuals
Involved
Process
22. Application of the CFIR
• Consists of 39 individual constructs
• Cannot use them all in every study
– And not all will apply
– A priori assessment of which constructs to include
• Modifiable & non-modifiable constructs
• Determine levels at which each construct may
apply
– E.g., teams, departments, clinics, regions
22
24. Selected Models for Discussion
3. PARiHS Framework3 Positive influence
• major domains
– Evidence High
– Context
– Facilitation Evidence
•Continuums of high and
High
n
t io
ta
low values that interrelate
c ili
Fa
Low Context
to influence
High
implementation Negative influence
Kitson A, Harvey G, McCormack B. Enabling the
implementation of evidence based practice: a
conceptual framework. Qual. Health Care
25. PARiHS-Diagnostic and Evaluation Grid
Figure 1
The PARiHS Diagnostic and
Evaluative Grid.
Evaluating the successful
implementation of evidence into
practice using the PARiHS
framework: theoretical and
practical challenges
Implement Sci. Implement
Sci;3:1-1.
26. Evidence Sub-elements:
Research evidence
Weak: Anecdotal evidence, descriptive
Strong: RCTs, evidence-based guidelines
Clinical experience.
Weak: Expert opinion divided
Strong: Consensus
Patient preferences and experiences
Weak: Patients not involved
Strong : Partnership with patients
Local information
28. Facilitation Sub-elements:
Characteristics (of the facilitator)
Weak: Low respect, credibility, empathy
Strong: High respect, credibility, empathy
Role.
Weak: Lack of role clarity
Strong: Clear roles
Style.
Weak: Inflexible, sporadic
Strong: Flexible, consistent
29. PARiHS Framework
Successful implementation is most likely to
occur when:
1.Scientific evidence is viewed as sound and
fitting with professional and patient beliefs.
2.The healthcare context is receptive to
implementation in terms of supportive
leadership, culture, and evaluative systems.
3.There are appropriate mechanisms in place to
facilitate implementation.
32. Theoretical or Conceptual
Framework
– More specific and concrete than theory
– Can usually be shown in a diagram/picture
– “… could be populated by multiple theories, at
multiple levels” (Kitson et al, 2008)
CIPRS: Stetler &
Damschroder Theoretical
Frameworks
33. Model
– Represents a specific situation
– Narrower in scope
– More precise in their assumptions – including
relationships (Kitson et a, 2008)
– May be used interchangeably with “framework”
(Sales, et al., 2006)
CIPRS: Stetler &
Damschroder Theoretical
Frameworks
34. Why NOT Theory?
• Just “use common sense”1
– Using theory doesn’t make it any easier to judge applicability of
evidence
– It isn’t clear how to translate theory reliably to study design
– So many theories why should any one of them be given
supremacy?
• BUT: “Common Sense” alone hasn’t worked so far…
BUT
– Trial & error approach
– Reinventing the wheel
– Cherry-picking interventions
– Retrospectively trying to understand the black box
CIPRS: Stetler & 1. Bhattacharyya O, Reeves S, Garfinkel S, Zwarenstein M. Designing theoretically-informed implementation
Damschroder Theoretical interventions: fine in theory, but evidence of effectiveness in practice is needed. Implement Sci 2006;1:5.
Frameworks
35. Why “Theory”: Dual Objectives
1. Generalize knowledge about how to implement and
sustain interventions
– Facilitate systematic accumulation of generalizable
knowledge
• Across studies
• Across settings
• Across interventions
• …other salient characteristics/factors
2. Replicate successful implementation
• Help navigate complexity of implementation &
sustainability
• Tailor essential factors to fit the context
From Damschroeder and Stetler 2011
36. Selecting a Theory - 1
• Consider Context
– Study characteristics
– Professional discipline/perspective
– Intervention characteristics
– Inner and outer setting
– Individuals involved
– Implementation process
• Consider Level
– Individuals
– Teams
– Organization
– System
CIPRS: Stetler &
Damschroder Theoretical
Frameworks
37. Selecting a Theory - 2
• Multiple theories often needed
– Process theories
• How implementation should be planned, organized and
scheduled
– Impact theories
• Hypotheses and assumptions about how
implementation activities will facilitate a desired
change, as well as the facilitators and barriers for
success
CIPRS: Stetler & Adapted from: Grol RP, Bosch MC, Hulscher ME, Eccles MP, Wensing M. Planning and studying
Damschroder Theoretical improvement in patient care: the use of theoretical perspectives. Milbank Q. 2007;85(1):93-138.
Frameworks
38. Selecting & Using a
Theoretical Framework
CIPRS:
Stetler/Damschroder,
Theoretical Frameworks
39. Uses/Potential Value [PDI*]
• Provide a way of thinking about a • Describe relationships among
study/project/implementation elements/constructs
• Focus the user on what is important to the • Guide development of hypotheses to test
issue implementation science
• Understand your EB-innovation/ • Identify concepts that may be of
recommendation/change importance and need to be statistically
• Develop a plan for formative evaluation; controlled or tracked
e.g., diagnostic analysis of barriers to and • Help with measurement
influences on using targeted best practices • Facilitate interpretation re: influences and
and applying an implementation strategy meanings
• Select and tailor interventions to promote • Identify boundaries around the
the use of evidence [Intervention mapping] project/study
• Assist with operational definitions of • Provide a framework for summarizing,
intervention element reporting findings
CIPRS: Stetler &
Damschroder Theoretical
Frameworks
40. Specific Selection and Use: QUERI
Centers
• Use:
– Facilitate strategic planning overall
– Guide implementation science goals
• Study, understand, predict causal
mechanisms/paths
• Evaluate chosen framework/s
• Selection:
– Open selection; one or multiple
• “NO NEW THEORIES” [Banff, International Conference,
CIPRS: Stetler &
2008]
Damschroder Theoretical
Frameworks
41. “Theoretical” Selection
• Origins of the framework
• Meaning of the “QUERI” Evaluation:
framework • In-depth understanding
• Logical consistency
• Theory criteria
• Generalizability
• Parsimony • Overall strengths
• Testability • Overall limitations
• Usefulness • Missing elements
[QUERI PDI Working Group]
[Grol et al., 2007]
CIPRS: Stetler &
Damschroder Theoretical
Frameworks
42. Specific Selection and Use: QUERI
Projects
• Use:
– Assessment
– Intervention planning
– Hypothesis generation
– Evaluation
• Black box of implementation and progressive/interim outcomes
• Usefulness of chosen theories
• Selection:
– Based on the issue at hand
– Apparent relevance of your “broad” center framework
• Strengths, limitations, relevance
CIPRS: Stetler &
Damschroder Theoretical
Frameworks
43. Project: Theory Selection and Use
1. Your targeted EBP recommendation:
a. Explore others’ relevant experiences and results (Grey & published
literature):
Did they use process/impact theories? Which and to what effect?
Evidence of WHY and HOW a particular intervention/ strategy did or did not
work therein?
Did they use isolated, atheoretical interventions or a multi-faceted
strategy?
Evidence of WHY and HOW a particular intervention did or did not work
therein?
What related barriers, facilitators, determinants have been identified?
Outside of QUERI?
Prior Center work, including Step 3 activity?
CIPRS: Stetler &
Damschroder Theoretical
Frameworks
44. Project: Theory Selection and Use cont’d
b. Understand the nature of your innovation (e.g., per Rogers):
What are its attributes/characteristics?
• “Core/peripheral” to the clinician’s sense of their practice
• Complex or “simple”
• Obvious appeal or the reverse; etc.
What are the potential targets of change?
• Per Level/s: Individual, team, clinic, organization? [CAVEAT:
“Individuals” work in a context]
• Per Stage/focus of change: Awareness, knowledge, skills, self-
perception, attitude, behavior, systems, structures, etc.
Given this information, have potentially influential factors been
clearly identified?
CIPRS: Stetler &
Damschroder Theoretical
Frameworks
45. Project: Theory Selection and Use cont’d
2. Choose “2” implementation theories to “try” as a “way of
thinking” about your particular issue:
– Per your general knowledge of their focus and or prior use
– Per their strengths, limitations, potential usefulness for your issue
2. Assess “fit” of these various frameworks and make
selection/s of one or more, as appropriate
CIPRS: Stetler &
Damschroder Theoretical
Frameworks
46. 4. Plan your implementation project in context of
the selected framework
a. Define relevant terms
• Conceptually
• Operationally /measurement and/or actions
b. Develop formative evaluation questions/tools [E.g., if PARIHS-
related]
• Local diagnostic analysis: E.g., assess likely barriers
(How do stakeholders perceive the attributes of the expected change?)
• Implementation-focused: E.g., actual barriers (To what extent does leadership
actually support the new practice or adoption efforts?)
• Progress-focused: E.g., interim staff performance on the new, expected innovation
(Relates to designated outcomes or “successful implementation”)
• Theoretical/Interpretive: E.g., to what extent did the framework provide an adequate
description of results and related influential factors? (Were any significant factors
missing?)
CIPRS: Stetler &
Damschroder Theoretical
Frameworks
47. 4. Plan… cont’d
c. Select interventions, per theory/theories and in light
of:
• Expected barriers
• Prior studies’ findings
• Local diagnostic data
(Simple example: Complex E-B delivery system)
• Per PARIHS model, consider use of an external facilitator & routine audit/feedback
system; based on identified concerns of clinicians regarding strength of the evidence,
additionally use social marketing actions/theory; based on prior similar study findings,
use other “facilitation” techniques such as clinical reminder.
d. Identify “theory-related” hypotheses to be tested
(Simple example: Complex E-B delivery system)
• Test hypothesis that sites with an external facilitator will be more successful than
comparison sites under analogous conditions of limited resources and passive
leadership.
CIPRS: Stetler &
Damschroder Theoretical
Frameworks
48. Challenges of Understanding the Black
Box
• Implementation theories under-studied
– May be “borrowed”
– Few have been critically analyzed for
strengths/limitations
• Measurement tools limited
• Published studies use of theory often unclear
or absent
CIPRS: Stetler &
Damschroder Theoretical
Frameworks
Notas del editor
From a public health perspective guidelines for practice cannot address realities of challenges for health faced in community settings And geographically, from a global health perspective, evidence is not implemented at all, implementation is not assessed for efficacy outcomes, and efforts are lacking in providing real integration into low cost effective strategies
Predisposing, Reinforcing, and Enabling Constructs in Educational/ environmental Diagnosis and Evaluation.
F 1 = facilitation method for transforming weak context and strong evidence into a highly receptive context2. F 2 = facilitation method to manage weak context and weak evidence situation – most challenging and possibly involves issues of safety, basic competence needs to be managed3. F 3 = facilitation method to manage strong context and weak evidence situation – issues of routine and power involved Proposed evaluative approach based upon PARiHS framework. In essence, the setting for a proposed implementation project can be evaluated in terms of the two broad elements of context and evidence, judged as falling somewhere on a weak to strong continuum according to degree of support for implementation. In terms of the four cells you see depicted here, then, you might have a setting with F1: context that is weak in support of implementation, but evidence is strong. This might be a unit with that is not organized for training and learning and it may have low morale, but the practice to be implemented may have strong empirical support that matches clinical experience and patient preferences well. F3: Setting for implementation in which Evidence is weak but Context is strong in support of implementation. This might be a unit with high morale and energizing, enabling leadership, but the practice to be implemented has limited empirical literature support and among clinicians opinions are highly divided on its utility. These two hypothetical scenarios would require different facilitation strategies F1: Might require facilitation strategies aiming to move the weak context elements to strong, e.g., develop training programs and foster collaborative relationships among staff F2: Might require facilitation strategies, such as interventions to develop professional consensus on best practices, or education on recommended best practices in other settings.
Research: Empirical research Weak support for implementation would consist of anecdotes, lore, descriptive studies Strong support RCTs, evidence-based guidelines Clinical: Experiences reflected upon and tested by individuals and groups Weak support for implementation would be conditions with absent or divide expert opinion. Strong support would be consensus across experts. Patient: Weak support for implementation would be conditions in which patient input was not included or patients would be opposed to implementation of the practice Strong support for implementation would be conditions where patient opinion is incorporated into the implementation or where patient preferences otherwise matched the practice to be implemented. Local: audits, local research, pilot projects. Weak support: no local experience or experiences are not consistent with the practice to be implemented Strong support: Local information matches the practice to be implemented.
Culture: Attitudes and beliefs about health care, organizational morale. Weak support for implementation are task driven settings with low moral Strong support for implementation is learning, innovation, and patient-centered approaches are valued Leadership Weak support for implementation are those with poor organizational structure, unclear roles Strong support for implementation possess effective organizational structure, clear roles Evaluation Strong support for implementation is found in settings with routine evaluation and feedback.
Facilitation refers to various strategies used to help change occur within an organization. Broadly classed as practical, problem-solving interventions, and supportive activities aimed at enhancing individuals and teams motivation and ability to implement a practice. Can be many things, but PARiHS emphasizes facilitation as characteristics of a person in the specific role of the Facilitator. Characteristics of facilitator Strong support for implementation is found in a facilitator who elicits respect, is credible, and empathic Role Strong support for implementation occurs when the role of the facilitator is clear and adequate for the implementation project. Style Personal styles characterized by flexibility and consistency give strong support for implementation.
To keep my thinking about the PARiHS framework organized I like to think of the three elements in the following narrative format. Summarizing the model with its three elements.