Applying behavioural science theories and frameworks in implementation science

Dr. Fabiana Lorencatto

Research Lead, UCL Centre for Behaviour Change, UK

4th BeSP Sypmosium

March 2020

@UCLBehaveChange@fabilorencatto

This talk

• Why behaviour change in implementation?

• Limitations typical approaches to implementation interventions

• Applying behavioural science• Frameworks and theories for systematic intervention design• Behaviour Change Wheel Approach

• Examples from healthcare context

Why behaviour change?

Implementation is challenging….

New intervention, practice, tool,

technology

Tell people about it

(e.g. guidelines, dissemination)

Uptake and implementation

The problem – evidence practice gaps:

• Despite training and guidelines, many do not always act in line with evidence-based recommendations

• Research• Netherlands: 30-40% of patients did not receive ‘evidence-based’ health care Grol et

al, 2001

• US: 20-25% received care that was unnecessary or even harmful Schuster et al, 2005

Implementation as behaviour change• Guidelines do not implement themselves!

• Implementation almost always requires someone to do act or do something differently

• E.g.

• These actions are all forms of human behaviour

Implementation as behaviour change • Implementation depends on changing behaviour of many different

types of people and roles at different levels in organisations, networks and systems

• Professionals, • support staff, • commissioners,• managers, • policy makers, • patients, etc.

Implementation interventions • ‘specific methods or techniques used to enhance the adoption, implementation, and

sustainability of a clinical program or practice’ (Proctor et al. 2013)

Many have achieved modest and

variable success

• Why?

• Often not approaching

implementation interventions

in terms of behaviour

change…

EPOC taxonomy

ISLAGIATT principle

It Seemed Like A Good Idea At

The Time

Prof. Martin Eccles, implementation researcher, UK

‘Hunches…Common sense’

lack rationale…

‘just educate’ ‘it worked elsewhere’‘guidelines’

Traditional approaches to designing interventions:

Sometimes ISLAGIATT works…more often it doesn’t….

Behavioural science• Disciplines dedicated to scientifically studying human behaviour:

• Offer theories, frameworks, methods for understanding what drives behaviour and how to use this to most effectively change it

Warning: no magic bullets or

universal truths

• Interrelated theories, frameworks and evidence-based principles

• Theoretical basis for linking barriers/facilitators to intervention strategies

• Guide decision making and facilitate systematic, step-by step, transparentand more effective approach to intervention design

• Policy makers, practitioners, researchers from different disciplines and levels of experience

Michie S, Van Stralen MM, West R. The behaviour change wheel: a new method for characterising

and designing behaviour change interventions. Implementation science. 2011 Dec;6(1):42.

Michie S, Atkins L, West R. The behaviour change wheel. A guide to designing

interventions. 1st ed. Great Britain: Silverback Publishing. 2014.

https://www.ucl.ac.uk/behaviour-change/files/bcw-summary.pdf

Behaviour Change Wheel approach

The Behaviour Change Wheel approach: key steps

What behaviour are you trying to change?

What will it take to bring about the desired change?

What types of broad intervention approaches

might be relevant?

What specifically components should my

intervention involve?

• Define ‘problem’ in behavioural terms• Map out system of behaviours• Who, needs to do what, when, where?

• ‘Behavioural Diagnosis’• Understand behaviour in context• Identify barriers/enablers to change

• Consider range of intervention strategies• Match choice to behavioural diagnosis

Step 1: Defining the problem in behavioural terms

• Tendency to think in terms of outcomes

• ‘reduce infection rates’ or ‘improve infection control’ ≠ a behaviour

• Product of numerous discrete behaviours

Hand-washing

Effective use of protective

clothing

Cleaning surfaces

Appropriate prescribing of treatments for infections

Antibiotic prescribingIdentifying and

diagnosing suspected infections

Communicating/

documenting/ escalating

Prescribing antibiotic + decision around dose, route, administration

Timely review of

antibiotic prescription

Stop/ de-escalate

Performed by:

• Multiple actors (care assistants, nurses, doctors, patients, family members)

• At different time points in care pathway

• In different settings

• Vary by type of patient and/or type of infection

Why does this matter?

• Influences will vary across different:• Behaviours, actors, settings, time, contexts

• Start by mapping out system• Do not need to intervene on full system! Focus on one aspect • Consider:

• Impact• Likelihood of change• Spillover

And therefore so might the type of intervention(s)

needed

Be specific: AACTT principle:• Asks: precisely who would do what, differently, to whom, when and where

that would lead to improved outcome?

Actor- who?

Action – needs to do what?

Context- where?

Timeframe- when?

Target- to whom?

Presseau J, McCleary N, Lorencatto F, Patey, A, Grimshaw J, & Francis JJ. (2019). Action, Actor, Context, Target, Time (AACTT): A framework for specifying behaviour. Implementation Sci

‘hand hygiene in hosptials’

Duncan et al. "A behavioural approach to specifying interventions: what insights can be gained for the reporting and implementation of interventions to reduce antibiotic use in hospitals?." Journal of Antimicrobial Chemotherapy (2020).

A (more!) behaviourally specific example:

Nurses (who) in intensive care (where)

should clean there hands with alcohol rub

(what) before and after (when) physical

contact with patients (whom) (Sun et al 2011)

Step 2. Understand the behaviour in context (‘Behavioural diagnosis’)

• Why are behaviours as they are?

• What would it take to implement?

• What would facilitate? What would hinder?

• Clinical practice is a form of human behaviour…

• Answering these Qs helped by a theory of

behaviour change

Prof. Susan Michie

Many theories applied to implementation!

Nilsen P. Making sense of implementation theories, models and frameworks. Implementation science. 2015

Dec;10(1):53.

Michie SF, West R, Campbell R, Brown J, Gainforth H. ABC of behaviour change theories. Silverback Publishing; 2014.

The COM-B system: Behaviour occurs as an interaction between three necessary conditions

Michie et al (2011) Implementation Science

Psychological AND Physical ability to enact the behavior Knowledge, Memory attention decision making, physical and social skills

Reflective AND Automatic mechanisms that activate or inhibit the behaviourIntentions, Goals, Perceived relevance, identity, Beliefs about consequences, Self-confidence, Rewards, incentives, sanctions, Emotions

Physical AND Social environmental factors that enables or inhibits the behaviourAccess, layout, resources, prompts, cuesSocial influences (pressure, support)

Theoretical Domains Framework

1. Knowledge2. Skills3. Memory, Attention, Decision

Making 4. Behavioural regulation

5. Social Influences6. Environmental context and resources

7. Emotions8. Social professional role and identity9. Goals10. Beliefs about consequences11. Reinforcement 12. Beliefs about capabilities13. Intentions14. Optimism

Michie et al. 2005 BMJ Qual & SafetyCane et al 2012 Implementation Sci

Am I aware of what I need to do (guidelines/evidence)?

Appropriate skills/training?How do I decide to do X?

Do you ever forget to do X?

Is doing X part of my clinical role? What will happen if I do X? What if I

don’t do X?Is it a priority?

How confident am I? How worried/ concerned?

Do I have sufficient resources (time/ staff/ equipment) to do X?

Is doing X influenced by my peers, managers, other professional

groups, patients, relatives?

Behavioural Diagnosis:

• Complex behavior

• Complex set of interacting influences…

• …beyond knowledge

• Need to consider broad range of intervention strategies…

• …beyond education!

• Match choice of strategy based on behaviouraldiagnosis

How do we move from diagnosis to intervention?

Facilitated by using a framework:• To outline/define variety of options and aid selection

• Several such frameworks to date…• Review of 19 frameworks related to health promotion, environment, culture change, social marketing

etc• Including EPOC taxonomy implementation strategies

• Overlapping, none comprehensive• None linked to model of behaviour change• Different assumptions/emphasis around what drives behaviour (attitudes, unconscious biases, social

environment)

Michie et al (2011) Implementation Science.

• So developed a synthesis of the 19 frameworks

Use rules to reduce the opportunity to engage in the behaviour (or to increase behaviour by reducing opportunity to engage in competing behaviours)

Increase knowledge or understanding

Use communication to induce positive or negative feelings to

stimulate action

Create an expectation of reward

Create an expectation of punishment or cost

Impart skills

Increase means or reduce barriers to increase

capability (beyond education or training) or opportunity

(beyond environmental restructuring)

Provide an example for people to aspire to or emulate

Change the physical or social context

But how will you ‘educate,’ ‘enable,’ ‘restrict,’ ‘model…’?• Intervention functions are made up of component Behaviour change techniques (BCTs)

• ‘active ingredients’ (e.g. goal setting, feedback, action planning)

BCT Taxonomy v1: 93 BCTs in 16 groupings (Michie et al. 2013)

Matching intervention strategy to diagnosisTo change …. Consider one or more of ….

Education – Training (i.e. build knowledge and

skills)

Persuasion - Incentivisation - Coercion -

Modelling (i.e. increase motivation to engage

in target behaviour; reduce motivation

competing behaviours)

Restriction- Environmental Restructuring –

Enablement (i.e. to increase opportunity target

beh/ reduce competing behaviours)

Selecting relevant functions: mapping tools

Intervention functions

Education Persuasion Incentivisation Coercion Training Restriction Environmental restructuring

Modelling Enablement

Physical capability

Psychological capability

Physical opportunity

Social opportunity

Automatic motivation

Reflective motivation

Based on expert consensus

Which BCTs? Mapping matrices

Michie et al. 2014

Narrowing down the list: APEASE criteria

Affordability Can it be delivered within an acceptable budget?

Practicability Can it be delivered as designed and to scale?

Effectiveness/

cost-effectivenessHow well does it work and is it worth the cost?

Acceptability Is it judged appropriate to relevant stakeholders (policy makers,

practitioners, the public) and engaging for potential users?

Side-effects/safety Does it have unwanted side-effects or unintended consequences?

Equity Will it reduce or increase disparities in health/wellbeing/standard of

living?

Michie et al. 2014

An example of intervention design using BCW

Multimorbidity = presence of 2+ conditions

Affects 60% patients primary care

Few interventions to improve medication management

1. Define the problem in behavioural terms

2. Select target behaviour

3. Specify the target behaviour

4. Identify what needs to change

Identify:

5. Intervention functions

6. Policy categories

Identify:

7. Behaviour change techniques

8. Mode of delivery

Stage 3: Identify content and implementation options

Stage 2: Identify intervention options

Stage 1: Understand the behaviour

What data did they draw on?

• Published literature

o Available evidence and reviews of medication management in multimorbidity

o Synthesised in a systematic review

• Primary research

o Interviews with 20 GPs on their approaches to prescribing

o Case focused

o Further define problem + selection target

Impact: likely large. Needs to happen first

for other behaviours to occur. Expressed in

all interviews.

Likelihood of change: High based on

interview findings. Easy to measure.

Spillover: High. Promote improvements in

other prescribing activities

Who? GPs

What? Active, purposeful medication review instead of passive ‘maintaining the status quo’

When? [unclear]

Where? In routine general practice

How often? Regular intervals (according to guidelines)

With whom? Patients with multimorbidity

1. Define the problem in behavioural terms

2. Select target behaviour

3. Specify the target behaviour

4. Identify what needs to change

Identify:

5. Intervention functions

6. Policy categories

Identify:

7. Behaviour change techniques

8. Mode of delivery

Stage 3: Identify content and implementation options

Stage 2: Identify intervention options

Stage 1: Understand the behaviour

How did they identify what needs to change?

• Used the COM-B model to identify GPs’ Capabilities, Opportunities and Motivations for engaging, or not

engaging, in active medication review

o Synthesised barriers and facilitators from the literature and their interviews with GPs

1. Define the problem in behavioural terms

2. Select target behaviour

3. Specify the target behaviour

4. Identify what needs to change

Identify:

5. Intervention functions

6. Policy categories

Identify:

7. Behaviour change techniques

8. Mode of delivery

Stage 3: Identify content and implementation options

Stage 2: Identify intervention options

Stage 1: Understand the behaviour

• Uncertain which medications were the most effective in the

presence of other conditions (psychological capability)

• GPs’ sense of isolation in the management of multimorbid patients(social opportunity)

• Lack of time to review medication during the consultation (physical opportunity)

• Automatically print off prescriptions, reviewing is not routine. Habit of ‘not rocking the boat’ (automatic motivation)

• Perception there are more important things to do in a consultation than review medications (reflective motivation)

1. Define the problem in behavioural terms

2. Select target behaviour

3. Specify the target behaviour

4. Identify what needs to change

Identify:

5. Intervention functions

6. Policy categories

Identify:

7. Behaviour change techniques

8. Mode of delivery

Stage 3: Identify content and implementation options

Stage 2: Identify intervention options

Stage 1: Understand the behaviour

How did they select which intervention functions?

1. Used the COM-B / intervention function matrix to identify potential

intervention functions– All intervention functions were potentially relevant

2. Used the APEASE criteria to select the most appropriate

intervention functions

1. Define the problem in behavioural terms

2. Select target behaviour

3. Specify the target behaviour

4. Identify what needs to change

Identify:

5. Intervention functions

6. Policy categories

Identify:

7. Behaviour change techniques

8. Mode of delivery

Stage 3: Identify content and implementation options

Stage 2: Identify intervention options

Stage 1: Understand the behaviour

How did they select which behaviour change techniques?

1. Drew up a long list from BCTTv1 of the most frequently used BCTs linked to enablement, incentives

and environmental restructuring. This resulted in 32 BCTs.

2. Expert panel rated each according to the APEASE criteria + in light of behavioural diagnosis

findings.

3. Selected 5 BCTs to include in intervention + panel agreed how to operationalize BCTS

End to end development reported in the paper

COM-B component

Barrier Intervention function/ technique

Example

CAPABILITY ‘uncertainty about what medications were most effective in the presence of other conditions’

‘professional isolation’

• F: Enablement• BCT: Social support (practical)• Action planning

Two GPs support each other to review medication (Social support practical)

GPs choose a day/time/office that suits them best and decide how many patient cases to review in one sitting (Action planning)

OPPORTUNITY ‘insufficient consultation time to review medications’

• F: Environmental restructuring• BCTs: Restructuring social

environment• prompts/cues

Agree on a policy of protected time to review medication (~1 hr p/week)

Use a standardised checklist of questions to cue the review

MOTIVATION ‘more important things to do’

• F: Incentives • BCT: self-incentives

GPs award themselves continuing professional development points: one point for each cumulative hour

Systematic, transparent, step-wise-approach for choice of intervention strategy

Australia• Management for Borderline

Gestational Diabetes Mellitus• Reduce hip/knee surgery in

osteoarthritis

Ghana• Reduce risk of infection in

changing toilet cartridges in slums

India• Smartphone app to reduce

cardiovascular disease risk

International Red Cross• Train volunteers

Kenya• Improve paediatric health care

Kosovo• Improve maternal health

guideline implementationNetherlands• An organisational

intervention tool

Papua New Guinea

• Change Betel nut chewing behaviour

Thailand• Preventing melioidosis

USA• Improving colorectal cancer

screening • Providing long-acting

reversible contraception to adolescents

• Improve parenting practices for children with challenging behaviour

International applications of the BCW

Summary and recommendations Implementation in healthcare requires behaviour change- but this is rarely considered when designing implementation interventions

Where possible, avoid ‘rushing’ to intervention and ISLIGIATT

Behavioural science has tools and theoretical frameworks to facilitate

Be specific about what (and who/where/when/whom) you wish to change.

Ask why would people change? What is driving behaviour?

Look beyond education: consider full range of intervention strategies and match to behavioural diagnosis

Be transparent and systematic: to maximise cumulative learning and aid implementation

Thank you for listening!

F.lorencatto@ucl.ac.uk

@UCLBehaveChange

@Fabilorencatto

CBC Summer Schools, London 2020July/August

Principles and Practice of

behavior change

Advanced Summer School

Motivation, Systems Mapping,

https://www.ucl.ac.uk/behaviour-change/training/summer-school

All proceeds from CBC teaching, training, books and

products go to further development

www.ucl.ac.uk/behaviour-change

https://www.futurelearn.com/courses/

behaviour-change