Skip main navigation

Four end to end examples of COM-B behavioural diagnosis applied to AMS

In this video, we read about various studies that showed how behavioural science has been applied to understand problems in antimicrobial stewardship.
Hello again. This is Fabiana Lorencatto. In this final module, I’m going to talk you through some end-to-end examples to illustrate how we can apply these behavioural science frameworks to design antimicrobial stewardship interventions. Let’s start with step two, conducting a behavioural diagnosis in a scenario of designing or identifying new potential interventions. Remember, the behavioural diagnosis is about exploring barriers and enablers to the behaviour of interest. The COM-B model and the theoretical domains framework, which Magda explained in an earlier module, had been increasingly applied to explore what factors are influencing antibiotic prescribing across different care settings.
I want to start by signposting you to some examples of studies that have used this approach to look at barriers and enablers to antibiotic stewardship in care homes, general dental practises, community, pharmacy, and hospital settings, using a mixture of qualitative interviews, focus groups, and/or surveys. These are screenshots of some papers which you might wish to read. In the next few slides, I will talk you through some examples of interview questions used in these studies and give you a flavour of some of their findings from these and a few other similar studies that have also used these frameworks to conduct behavioural diagnoses.
However, if this is of interest I would highly recommend reading these papers for further detail, which include more information on what they did, what they found, and the discussions and implications of these findings. All of these papers are freely available through Open Access. Let’s start with the questions asked during the behavioural diagnosis phase of these qualitative interviews studies. Here are some examples of questions asked to tap into the role of capability, opportunity, and motivation in terms of how they influence antibiotic stewardship across different care settings. Questions to explore capability, which, as you recall, can be physical or psychological– so knowledge and skills–
can include questions around: understanding of AMR, what antibiotics can and cannot treat, use of and awareness of guidelines for antibiotic prescribing, procedures for diagnosing, and treating infections.
Questions to explore motivation, which includes things like beliefs about consequences, goals, competing priority, emotions, et cetera, can include asking someone about the consequences or what they think will happen if they do prescribe or don’t prescribe an antibiotic. Will it make a difference to antimicrobial resistance? Do they have any concerns or fears or anxieties around prescribing an antibiotic or not? Is promoting antimicrobial stewardship something they see as part of their role? Are there any clinical competing priorities, and do they have any incentives in place to use antibiotics less?
And lastly, questions to explore physical and social opportunity include firstly, in terms of social opportunity, asking about team working dynamics and communication, role clarity, patient pressure, et cetera. Questions around physical opportunity can include availability of diagnostic tests, different types of antibiotics, time, costs, et cetera. So what do these studies find? Here’s an example of findings that were coded under capability. Well, first, it’s important to note that mostly the barriers identified related to psychological capability– so knowledge, rather than physical capability and skills. It appears that psychological capability played a mixed role in influencing antimicrobial stewardship.
While most healthcare professionals across care settings were aware of antimicrobial resistance, knew what it was, and that it was a growing public health threat, many reported having poor clinical microbiology knowledge and had incorrect or poor dose-adjustment skills. There was variable knowledge of guidelines, with some reporting that they knew the existence of and were familiar with the content of guidelines, both locally and nationally, whereas others had not even heard of the guidelines. There was also a lack of knowledge and skill around diagnosis and management of infection for particular patient groups, such as those with comorbidity or the elderly.
Lastly, some lacked knowledge of their specific local context, such as how and to what extent antibiotics were being used appropriately in their care home, practise, ward, or professional group. When it came to motivation, many of the barriers identified were emotional in nature due to a strong fear and anxiety of missing something if they did not prescribe an antibiotic and the potential harm to patient that could result. This was closely linked to the view that alternatives to not prescribing were just not as effective, so it’s best to prescribe antibiotics to address the infection in a timely manner.
Many reported mixed views as to whether engaging in antibiotic stewardship would actually make a difference to the threat of AMR or whether it was just a drop in the ocean that would not have any impact. Importantly, antimicrobial stewardship is relevant to almost every area of hospital practise, but for many, it was not their sole or key priority with many other competing clinical priorities and demands. So antibiotic stewardship fell much further down the list and did not get the attention it perhaps deserved. Many also felt confident in their own clinical experience and judgement and expertise, and this was seen to override guidelines that were perhaps put into place to try to encourage more prudent and appropriate use of antibiotics.
When it came to opportunity, this closely interacts with motivation. Concerning physical opportunity, much of this concerned the lack of equipment, timely interpretation of microbiology test results, high workload, and lack of time, all those key barriers to the opportunity to engage in stewardship. These interact closely with motivational barriers around competing priorities and clinical demands. High workload and competing priorities led to barriers around social opportunity, such as lack of role clarity as to who is responsible for antimicrobial stewardship and decision-making. In contexts such as surgery, responsibility for stewardship was seen as diffuse and uncoordinated with no clear line of responsibility.
There is also evidence in the broader literature around social opportunity barriers related to prescribing etiquette and hierarchies in which more junior health care professionals do not question the antimicrobial-prescribing decisions of more senior colleagues. Also, barriers within team dynamics are often identified, such as nurses acting as gatekeepers in care homes, identifying patients with potential infections, and communicating with general practitioners to get an antibiotic prescription. Sometimes, this was felt as a pressure by doctors and pharmacists. But this pressure to prescribe can also come from an external source. Patients themselves and their families were often cited as a strong barrier within social opportunity with patients sometimes demanding an antibiotic, despite potentially not needing one.
I hope that has given you a flavour of the types of barriers and enablers that can emerge within each domain of COM-B. It is not intended to be an exhaustive list and is just a set of examples. But having identified these influences, how can we take these findings from a COM-B-based behavioural diagnosis to move to intervention– that is, select the types of interventions that are going to be appropriate and effective in addressing barriers and enablers identified within each domain of COM-B? This is where the behaviour change wheel and behaviour change technique taxonomy come in.
Having done our behavioural diagnosis, we can now consult this matrix to identify the types of interventions that will target our barriers and enablers within capability, opportunity, and motivation. Remember, a blue cell indicates a match between that type of intervention strategy and the influence it is trying to address within COM-B. Many of those example studies I signposted to at the start of this module have also undergone this systematic and transparent intervention development process. They first did their behaviour diagnosis based on COM-B, or the theoretical domains framework to identify what needs to change. They then consulted the intervention mapping matrices and tables which pair COM-B and the behaviour change wheel and BCT taxonomy together to identify potential intervention strategies.
These papers include clearly described tables that set out and summarise the intervention development process in a stepwise manner starting with what they found in their behavioural diagnosis on the left, the domains from the models they pertain to– so either theoretical domains framework or COM-B– the intervention functions and more specific behaviour change techniques they selected, and recommendations and examples as to how that could be delivered in the local setting. This is an example of an intervention development table by a recent study by Jones et al., looking at antimicrobial stewardship in community pharmacy. There’s no need to read this table in detail at this stage. In the next slide, I will give you an example of an intervention for each COM-B domain.
I would, however, recommend reading over this paper and others for more information and detail. The key thing I want to emphasise at this stage is that this approach is the opposite of the “it seemed like a good idea at the time” ISLAGIATT black-box approach to intervention design I referred to in the first introductory module to this course. Instead, this is a clear, systematic, and transparent approach to communicating the intervention development process. It has a clear, evidence-based rationale as to why an intervention strategy was chosen– that is, what identified behavioural influence it is tackling, plus a clear specification of what precisely is in the intervention. It opens up the black box that are often behaviour change interventions.
In this table, I want to briefly give you examples of different types of interventions that could address barriers and enablers within capability, opportunity, and motivation to antimicrobial stewardship based on some of the example barriers and enablers identified in the earlier behavioural diagnosis studies. For instance, Chaves et al. identified in their study of hospital-based antimicrobial stewardship that some doctors reported poor microbiology knowledge and lacked skills around dose adjustment for certain medications. This is a barrier related to psychological capability, knowledge, and skills. You can increase knowledge and skills through education and enablement. Behaviour change techniques that support this are social support and instruction on how to perform the Behaviour.
They propose social support could involve having a pharmacist perform the dose adjustment to reduce errors, and further instruction on how to perform that Behaviour could be done by developing specialty-specific guidelines which clarify indications and also increasing overall undergraduate and continuing professional education in this area. Another example around motivation is from Fleming et al., whose study looked to antibiotic prescribing in care homes. They found many homes lacked clear goals and targets around appropriate antibiotic prescribing, which interacted with knowledge barriers around not knowing what current antibiotic prescribing practise was locally within their care home and lack of prescribing guidelines specific to elderly and comorbid patients. Potential intervention functions to address these motivational barriers include persuasion and enablement.
They propose functions such as these could be delivered through an audit-and-feedback-type intervention, which could involve goal setting, monitoring, feedback, and social comparison Behaviour change techniques. The intervention could involve setting goals in the form of guidelines and targets around appropriate antibiotic use specific for the care-home setting, followed by monitoring practise locally for a specified time period, then feeding this back to the care homes, outlining any deviations from guidelines and benchmarking against the set targets and other care homes locally. This can both enable care homes to change by creating an opportunity for them to reflect and to respond to deviations locally.
Highlighting discrepancies between current practise and a goal/standard can also motivate someone to change in order to try to reduce discrepancy– reach the goal. Motivation may be further increased by the comparison against others in terms of, first of all, showing that it can be done if other care homes are performing better or through friendly competition. In terms of opportunity in this hospital based study by Chaves et al., they identified physical opportunity barriers concerning the fact that contacting infectious disease physicians for antibiotic prescription approvals was far from straightforward and was not a time-efficient process. They propose this could be addressed by restructuring the physical environment through the introduction of a web-based decision support and automated approval system.
Regarding social opportunity, Charani et al.’s study of stewardship and surgical teams found strong barriers related to prescribing etiquette with lack of adherence to guidelines and a reluctance to challenge the practise of peers or more senior colleagues. Potential strategies to address this include modelling and enablement where more senior doctors within the specialty adopt a leadership and champion role, advocating the use of guidelines, and encouraging discussion and review of antibiotic prescribing practise and decisions in regular team meetings. These are just a couple of brief examples. Further examples of Behavioural diagnoses and potential intervention approaches are discussed and summarised in this Open Access paper by Peter, myself, and colleagues, which we would encourage you to read for further information.
Thus far, we have been discussing the design of interventions from scratch– that is, starting with a blank canvas and building the intervention incrementally. However, this blank canvas scenario is not always the case. Indeed, a more common scenario on health care quality improvement is that of having an intervention that has already been implemented in practise and perhaps is not working as well or as optimally as it could be. In this scenario, our goal might be to refine or build upon the existing intervention to try and improve its effectiveness. In this final section of the module, I want to talk you through an example of how these frameworks can be applied to refine an existing quality improvement intervention.
In this approach, we are basically working backwards through the process we have discussed so far. Our starting point is actually stage three, first of all, to characterise and describe what is in the existing intervention. What are the behaviours targeted, and what intervention functions and Behaviour change techniques are included in our existing intervention? We might then go to step two and conduct a behavioural diagnosis to explore why the intervention is not working optimally. What are the barriers and enablers to implementing the intervention, and does the intervention need to change in order to increase implementation?
We can then map the answers to these questions against one another to ask, first of all, to what extent do existing interventions target key influences on the Behaviour and barriers and enablers to implementation? Where are the gaps, and what additional strategies can we include to improve implementation and Behaviour change? The example I’m going to talk you through now is an intervention to improve the implementation of sepsis care bundle, the Sepsis Six. A multi-component Sepsis Six implementation intervention was designed through trial and error without the use of behavioural theory.
Content was not fully reported, but broadly, the intervention took the form of introductory group education and training, target setting, audit, group feedback of audit results, individual personalised feedback to staff involved in incidents when the bundle was not fully implemented, and environmental changes, including promotional documents. The intervention was initially implemented over a four-year period in three pilot wards of a large UK teaching hospital. The intervention was actually very successful at first. It initially achieved great success, increasing implementation of the bundle from 20% to 80%, and sepsis mortality dropped from 22% to 12%.
However, this took four years, and the 95% target was not reached for all pilot areas, raising the question of how best to develop the current intervention to achieve and sustain the 95% target before extending implementation elsewhere to the rest of the hospital. Through a multidisciplinary collaboration, Steinmo et al. applied the Behaviour change wheel, COM-B, theoretical domains framework, and Behaviour change technique taxonomy to try and refine the existing Sepsis Six implementation intervention and build upon it to improve effectiveness. Their first step was to describe what was in the existing intervention. They did this using a mixture of methods. To obtain information about what was in the intervention, the collected intervention documents, such as protocols, educational materials, et cetera.
They also conducted interviews with the people responsible for providing the intervention as well as those receiving the intervention and observed the intervention sessions being delivered. Based on the evidence they collected from these three sources, they coded and specified the content of the intervention in terms of the Behaviour change techniques and functions from the Behaviour change wheel and Behaviour change technique taxonomy. These behavioural science tools were used as frameworks to help them identify and classify what was in the interventions. This provided a baseline description as to what was in the existing intervention, and that is this table here from the paper, which summarises what BCTs were used in the existing intervention and what data sources if they were identified from.
Steinmo et al. then went on to explore the barriers and enablers to implementing the Sepsis Six care bundle. They did this by conducting semi-structured interviews with 34 hospital staff of different health care professional roles, including nurses, doctors, midwives, and health care assistants. They structured their interview questions around the COM-B and theoretical domains framework. Here are some examples of questions they asked to tap into each domain of COM-B during their behavioural diagnosis interviews. A full list of questions is available in the paper cited below. To assess capability, they asked, are you aware of the steps of the Sepsis Six care bundle? Can you talk me through these? Is remembering the steps of the Sepsis Six ever an issue?
How confident are you in your abilities and that of your colleagues to keep up with the Sepsis Six and implement it?
Questions to assess motivation, amongst others, included, how much difference do you think implementing the Sepsis Six makes? There are always competing priorities at work. Do you think implementing the Sepsis Six is a priority? Are there times when it is not a priority or becomes less of a priority? Do other things get in the way?
And lastly, to explore opportunity– do you have access to the necessary resources needed to implement the Sepsis Six? What extent do you see implementing the bundle part of your role? Are some steps more relevant than others? To what extent does working as a team facilitate or hinder implementation of the Sepsis Six? And here is a non-exhaustive list of some of the barriers and enablers they identified.
Within psychological capability, reported barriers included staff’s lack of familiarity with the steps in the sepsis six bundle– basically, not knowing what to do and when. Staff also reported lack of knowledge and confidence as to how and when to implement the bundle for more complicated scenarios and patients, for instance, what to do when there is no definitive diagnosis and whether to attribute symptoms to sepsis, when patients have been hospitalised for a while, and there’s lack of documentation as to when fluids or oxygen were started, and whether any of the other or earlier elements the bundles had already been administrated, so what to do in those scenarios where perhaps you don’t have the information you need.
In terms of motivation, barriers concern beliefs about consequences.
So what should happen or what would happen if the bundle was or was not implemented? Many reported fear of harming patients, not trusting or having confidence in the evidence underpinning the Sepsis Six care bundle– so whether it would actually work and benefit patients– and needing to be convinced that the benefits outweigh the risk and making sure they could see improvements in the patients in a timely manner.
Lastly, opportunity included many of the key barriers. The Sepsis Six bundle requires multiple actions potentially performed by different roles– individuals– at different time points. This resulted in lack of role clarity as to who should do what, where, and when and conflict around this. This also contributed to and was impacted by lack of communication and teamwork. There are also very practical issues around the physical environment. For instance, the equipment needed to perform all the steps in the bundle in a timely manner was not always readily available in one single place. Sometimes, it was even dispersed across different ends of the hospital.
This was a particular issue during night shifts when the necessary staff needed to perform certain steps were not always readily present. Steinmo et al. used the Behavioural diagnosis to critically review the existing intervention and identify areas where the existing intervention components could be refined or built upon by introducing new or additional intervention components to better address the identified barriers and enablers. They did this through stakeholder review of the findings and through consultation with the Behaviour change wheel and Behaviour change technique taxonomy. On this basis, they proposed 10 modifications to the existing intervention. This involved adding four additional intervention functions and 12 further Behaviour change techniques. These are listed in detail in the paper.
On the following slide, I will talk you through three examples related to different components of COM-B. For instance, to adjust barriers within capabilities, such as not knowing what to do and when, the team selected the intervention function education and the Behaviour change technique instruction on how to perform the Behaviour. They proposed this could be done by having an educational frequently asked question information sheet with detail around appropriate fluid volumes, evidence of oxygen administration, data on low C. diff for broad-spectrum antibiotics, information about when the hour begins, et cetera.
To address motivational barriers concerning lack of perceived benefit and effectiveness of the bundle and fear of patient harm, they propose to use a persuasive intervention strategy, involving information about the health consequences and feedback, for example, by providing staff with feedback and evidence on beneficial patient outcomes as a result of having received the care bundle. Lastly, to address barriers around physical opportunities, such as not having the necessary equipment to hand in one place, they propose the intervention function environmental restructuring and the Behaviour change technique adding objects to the environment in the form of introducing sepsis trolleys or bags that had all the needed equipment readily available in one place.
I hope this module has provided you with an overview of how this behavioural science approach and toolbox of frameworks can be used flexibly to address different quality improvement scenarios, from designing new interventions to refining existing ones. We have also signposted you to a number of examples which you can access for further information and detail as needed. Lastly, I just want to conclude this module with a final note a reminder to look beyond antimicrobial stewardship examples. There is a growing body of evidence in the behavioural and implementation science literature focused on understanding what features make for more effective quality improvement interventions.
For instance, audit and feedback is one of the most widely used quality improvement interventions, yet its effects are highly variable. There have been reviews to try to identify what makes for more effective audit and feedback. Specifically, what characteristics and modes of deliveries are associated with improved outcomes? These identify that feedback is more effective when it is delivered more than once in writing and in verbally by respected peer or colleague, rather than an external source or regulator, targets behaviours for which there are room for improvement, are accompanied by explicit recommendations and action plans for change, and some evidence for the use of multiple comparators, such as top 10% of peers, regional, national comparison, et cetera, and also is supportive, rather than punitive in tone.
For instance, if you decided to deliver an audit and feedback intervention to improve antimicrobial stewardship, this is one source of more general evidence that you could consult to help guide your decision-making around how best to design the intervention features to maximise its likely effectiveness. Thank you very much for listening.

In this video, Dr Fabiana Lorencatto discusses various studies which show how behavioural science has been applied to understand problems in antimicrobial stewardship and how it can be worked upon.

In Step 2.2, the Behaviour Change Wheel and COM-B approach was introduced and in this video we can see how it can be applied. AMR is a growing concern worldwide and using the COM-B approach helps us break down the problem into its various contributing factors.

Please find a downloadable PDF of the PowerPoint slides below.

This article is from the free online

Utilising Social Science and Behaviour Change in Antimicrobial Stewardship Programmes: Improving Healthcare

Created by
FutureLearn - Learning For Life

Our purpose is to transform access to education.

We offer a diverse selection of courses from leading universities and cultural institutions from around the world. These are delivered one step at a time, and are accessible on mobile, tablet and desktop, so you can fit learning around your life.

We believe learning should be an enjoyable, social experience, so our courses offer the opportunity to discuss what you’re learning with others as you go, helping you make fresh discoveries and form new ideas.
You can unlock new opportunities with unlimited access to hundreds of online short courses for a year by subscribing to our Unlimited package. Build your knowledge with top universities and organisations.

Learn more about how FutureLearn is transforming access to education