Moving from behavioural diagnosis to intervention

Hello. This is Fabiana Lorencatto again. In this module, I’m going to briefly introduce the concepts and frameworks that enable us to move from behavioural diagnosis to intervention. As a reminder, here are the three key steps in a behavioural science intervention development process. In earlier modules, myself and Jill explained the first step, precisely defining what is it that you wish to change in behavioural terms. Then Magda and Eilidh explained step two– how to conduct a behavioural diagnosis using theories and frameworks from the behavioural science to understand why behaviour is what it is and what would it take to bring about change, with examples of how we can apply these to explore factors influencing antimicrobial stewardship behaviours.

In this module, we are going to pick up where they left off and move on to the third and final step in the process. We will discuss how you can use behavioural science frameworks to effectively move from diagnosis to intervention by selecting the type of intervention, strategies, and components that are most likely to be relevant and effective in addressing the barriers and enablers identified in your behavioural diagnosis. To help you move from intervention to diagnosis, we will briefly introduce two additional behavioural science frameworks, the Behaviour Change Wheel and the Behaviour Change Technique Taxonomy. Both of these frameworks outline and define different types of ways and strategies through which we can change Behaviour.

I will explain how these can be used to guide decision making and intervention development in terms of deciding what the content of your interventions should be. That is, which types of Behaviour change strategies should I include that are most likely to be relevant and effective in addressing the barriers and enablers to my Behaviour of interest identified during my behavioural diagnosis or step two in the intervention development process. As a reminder, we discussed in earlier modules that conducting a behavioural diagnosis involves understanding Behaviour and context, that is, exploring what is influencing current Behaviour and what would facilitate or hinder change.

This can be done through just informal discussion with colleagues and key stakeholders or more formally by collecting research evidence through surveys, interviews, focus groups, literature reviews, et cetera. It is helpful to structure your behavioural diagnosis around a theory or model of Behaviour change, such as the COM-B model. Such models can highlight different types of influences on Behaviour, which you may wish to explore the role that they play in relation to your Behaviour of interest. For instance, what role does capability, opportunity, and motivation play in facilitating or hindering antimicrobial stewardship? Here is an example of some questions you might ask to explore this for each domain in your Behaviour diagnosis.

What you often find when conducting a behavioural diagnosis is that more often than not, Behaviour is unsurprisingly complex, in that it is influenced by a complex set of interacting individual, social-cultural, and environmental influences beyond knowledge deficits alone. Therefore, when it comes to selecting what type of intervention to use, it is important to start by considering the full broad range of potential interventions that may address these different types of influences beyond just education strategies alone. Considering the wide range of potential influences is facilitated by using a framework, which lists and describes the different types of interventions available, a menu of interventions of sorts. One such framework is the Behaviour Change Wheel.

The Behaviour Change Wheel was developed based on a systematic review of existing intervention frameworks in the literature. This review identified 19 different frameworks listing different types of interventions. Several things became apparent when reviewing the frameworks. First, these frameworks came from very different disciplines, and therefore, they had very different focuses. Some focused on describing interventions to change social norms, others on types of interventions to change the environment or individual motivation. As such, the review found that none of these existing 19 frameworks were comprehensive and that no single framework covered all of the other types of interventions listed in the other frameworks.

Also, none of these frameworks were linked to a model or theory of Behaviour change and therefore, proposed no basis for selecting interventions to address different types of influences and drivers of Behaviour. As such, the Behaviour Change Wheel was produced by synthesising or integrating the 19 different intervention frameworks into a single unified framework. This framework proposes nine different intervention functions or broad means by which we can change Behaviour. There are also seven policy categories, which are strategies for supporting implementation of interventions. In this module, we will further describe intervention functions but unfortunately, do not have scope to go into detail on policy categories. If this is of interest, I advise reading the paper cited here.

The Behaviour Change Wheel is linked to the COM-B model to propose which types of intervention functions are likely to be relevant and effective in addressing barriers and enablers within capability, opportunity, and motivation. This is the Behaviour Change Wheel. As you will see, at the hub of the wheel, in green, is the COM-B model, so the influences on the Behaviour of interest.

Around the perimeter, in red, are the nine different intervention functions or ways through which we can change Behaviour. The first of these is the well-known education or increasing knowledge or understanding, for instance, by providing information about the health consequences of inappropriate use of antibiotics. Then there is the increasingly well-known environmental restructuring. This involves introducing changes into the physical or social environment. These are what are often referred to as nudge-type interventions, which you may have heard of before.

For example, when healthier foods are arranged at the display so that they are in line with your vision to try and nudge you towards making healthier food choices, that’s a way in which we have changed the physical environment to try to change our food purchasing Behaviour. This function can involve making changes to the physical environment and materials, such as changing the availability of different resources or layouts, for example, introducing new antibiotics, changing the structure of patient notes to include checkboxes or fields related to antibiotic decision making and prescribing. Or it can also involve changes to the social environment in some way.

Such as changing team roles, dynamics, and composition, for instance, by introducing a new specialist antimicrobial stewardship pharmacist or nurse. The other functions include persuasion, which involves using communication and imagery to induce positive or negative feelings to stimulate action. This could include, for example, using posters with vivid imagery to convey the consequences and threat of AMR. Then there is incentivization, which involves creating an expectation of reward, which could involve material incentives, such as financial rewards for more appropriate antibiotic prescribing or also social incentives and rewards, such as praise and recognition for good practise. Conversely, there is coercion, which involves creating an expectation of punishment or cost.

Now, punishment is often unethical to do, so it’s not something we see very often in Behaviour change interventions. But an example of this is the threat of losing your driving licence through drink driving. Then there is training. Whilst education focuses on imparting knowledge, training focuses on building and imparting skills through practise and rehearsal of a desired skill or action, such as how to insert a catheter to minimise risk of infection. That involves training someone on the specific technical skill. There is then enablement. This is a bit of a catch-all category for ways to increase means and reduce barriers to increase capability and opportunity.

This could involve, for instance, offering someone practical and social support, helping someone to problem solve, action plan, monitor their Behaviour, giving someone feedback on their practise, et cetera. Then there is modelling, which involves providing an example for people to aspire to or emulate. This can involve having another physician demonstrate the desired Behaviour, such as reviewing an antibiotic prescription or hand hygiene, or using local champions. Lastly, there is restriction, which involves using rules to reduce the opportunity to engage in the Behaviour. This could involve, for instance, local hospital guidelines restricting or discouraging the use of certain antibiotics for certain symptoms or situations. Interventions almost always have more than one function.

For example, an intervention to increase hand hygiene could involve restructuring the physical environment, such as ensuring alcohol gel dispensers are available at the bedside of every patient; modelling, such as having senior nurses model and champion the use of alcohol rub pre and post patient contact; training on how to properly clean one’s hand using the alcohol rub, and education or persuasion around the health benefits of routine hand hygiene. These intervention functions represent the broad means through which we might change Behaviour. They are very much high-level categories or types of interventions. There is, therefore, a question as to how do we actually do these functions? That is, how will we educate? How will we enable, incentivize, model, restrict, et cetera?

Intervention functions can be broken down further into smaller component Behaviour change techniques. Behaviour change techniques are the smallest active ingredients of an intervention through which we bring about change. If you think of the intervention function as a cake, the Behaviour change techniques, or BCTs, as I’ll refer to them going forward, are the ingredients that go into the cake. The BCTs are the flour, the eggs, the sugar, et cetera. Example of Behaviour change techniques are things such as setting goals, action planning, providing rewards, feedback on Behaviour, demonstrating and Behaviour, providing information about health or environmental consequences, et cetera. There are many different types of Behaviour change techniques. In fact, behavioural science has a taxonomy of 93 different Behaviour change techniques.

This taxonomy aims to provide a common language through which to describe the content of Behaviour change interventions and to help guide the selection of intervention content or components. Each of the BCTs in the taxonomy is arranged hierarchically into 16 groups, representing the broad way through which these BCTs may change Behaviour. Each BCT has a label, a definition, and an example as demonstrated here for goal setting. This framework is therefore, essentially our encyclopaedia or periodic table of elements of different strategies and techniques through which we might change Behaviour. So there are lots of different ways we can change Behaviour. There are nine broad functions and 93 more specific Behaviour change techniques. How do we pick among these?

How do we use these frameworks to help us choose which types of interventions will address the key influences or determinants on the Behaviour I am targeting for change? Remember, the key principle underpinning this approach to intervention development is that of basing the choice of intervention approach on a theoretical understanding of what is influencing the Behaviour of interest, that is, a behavioural diagnosis. Interventions are much more likely to be effective if they include components that target the key influences on Behaviour.

Therefore, to facilitate this matching process, these behavioural science frameworks have essentially been mapped against one another to propose which functions from the Behaviour Change Wheel are likely to be more relevant and effective in addressing different influences within capability, opportunity, and motivation. This slide demonstrates the results of this mapping exercise. On the left, we have the domains of the COM-B model, on the right, the intervention functions from the Behaviour Change Wheel that are relevant to that domain. For example, capability, you will recall, is about knowledge and skills.

Therefore, if you find from your behavioural diagnosis that capability is your main barrier or enabler that you need to target for change, then you can increase capability through the functions of education and training to build knowledge and skills. Motivation can be increased through persuasion and modelling and also through the use of incentivization or coercion, that is, threat or punishment. These can increase motivation to engage in the desired Behaviour whilst also minimising any motivation to engage in any competing or undesired behaviours. Lastly, we can increase opportunity, which is all about how the physical and social environment facilitates or hinders our behaviours by restructuring the physical and social environment to bring about change in context.

For instance, we can do that through environmental restructuring, restriction, or enablement. This figure is just another way of depicting the content I explained on the previous slide. This is a matrix from the Behaviour Change Wheel guide, slash, book, which has the company domains along the left as the rows in the table. And the columns are the different intervention functions from the Behaviour Change Wheel. A cell shaded in blue indicates a match between that intervention function and COM-B domain, suggesting that that type of intervention strategy is likely to be appropriate and effective in addressing influences within that domain of COM-B. Therefore, you need to first conduct your Behavioural diagnosis to identify what is it that needs to change, which COM-B domain.

Then use the frameworks to identify which type of strategy to use. So for example, if you find from your behavioural diagnosis that reflective motivation is the main barrier to your Behaviour of interest, then you may wish to use either education, persuasion, incentivization, coercion, or modelling, the cells shaded in blue for reflective motivation as strategies to increase reflective motivation We have talked about how you select which intervention functions to use based on your behavioural diagnosis. Now, I want to briefly discuss which Behaviour change techniques, or BCTs, to include. Remember, intervention functions are made up of smaller BCTs from the 93-item BCT taxonomy.

In the Behaviour Change Wheel guide, there are intervention tables, which list the nine intervention functions and which BCTs in the taxonomy serve each individual function. Therefore, for instance, if you selected education as your intervention function, you can consult this table to identify which BCTs you may use to serve that function, so which BCTs you may use to educate. These could include, for example, providing information on social environmental consequences, health consequences, feedback on Behaviour, et cetera. It is important to acknowledge that these frameworks can only take you so far. As you have seen from my previous slide, there are multiple functions that you can pick from for each COM-B domain, and in turn, multiple BCTs serving each function.

Although interventions do typically include multiple functions and BCTs, it is unlikely that it will be feasible for your intervention to include all functions and all possible BCTs proposed in these tables. Furthermore, not all types of interventions will be feasible to do for your Behaviour of interest given your local context and available resources. You therefore, need to narrow it down and use local judgement to help you select from this long list of potential candidate interventions. One way you can do this is through consultation with colleagues and stakeholders to review each potential intervention strategy and technique against the APEASE criteria. Taking into account your Behaviour of interest, local context, and resources, ask yourselves, is this potential intervention affordable?

Can it be delivered within an acceptable budget? Is it likely to be practical to deliver? Can we deliver it as designed, and to scale? Is it likely to work? Is it likely to be effective and cost effective, so worth the cost? Is there any evidence in the literature that this type of intervention has worked before for this type of Behaviour and setting? Is it likely to be judged acceptable and appropriate to the relevant stakeholders, be that policymakers, practitioners, patients, the public, and also engaging and interesting for potential users or the people whose Behaviour we’re trying to change? Is it likely to have any side effects that are unintended or unwanted or any other negative consequences?

Is it something that’s safe to do? And lastly, is it equitable? Will it reduce or increase disparities in health, well-being, and standards of living? I hope this brief module has provided you with an introductory overview of these two additional frameworks– the Behaviour Change Wheel and Behaviour Change Technique Taxonomy. In the upcoming modules, we will talk you through how these can be applied to design and evaluate antimicrobial stewardship interventions, including identifying Behaviour change techniques and existing interventions, identifying the active ingredients or intervention functions in BCTs that are contributing to intervention effectiveness, and end-to-end examples of designing new and refining existing antimicrobial stewardship interventions using these frameworks.