Background
Development of the TDF
Version 1[15] | |
Domain | Constructs |
Knowledge | Knowledge Knowledge about condition/scientific rationale Schemas + mindsets + illness representations Procedural knowledge |
Skills | Skills Competence/ability/skill assessment Practice/skills development Interpersonal skills Coping strategies |
Social/professional role and identity | Identity Professional identity/boundaries/role Group/social identity Social/group norms Alienation/organisational commitment |
Beliefs about capabilities | Self-efficacy Control—of behaviour and material and Social environment Perceived competence Self-confidence/professional confidence Empowerment Self-esteem Perceived behavioural control Optimism/pessimism |
Beliefs about consequences | Outcome expectancies Anticipated regret Appraisal/evaluation/review Consequents Attitudes Contingencies Reinforcement/punishment/consequences Incentives/rewards Beliefs Unrealistic optimism Salient events/sensitisation/critical incidents Characteristics of outcome expectancies—physical, social, emotional; sanctions/rewards, proximal/distal, valued/not valued, probable/improbable, salient/not salient, perceived risk/threat |
Motivation and goals | Intention; stability of intention/certainty of intention Goals (autonomous, controlled) Goal target/setting Goal priority Intrinsic motivation Commitment Distal and proximal goals Transtheoretical model and stages of change |
Memory, attention and decision processes | Memory Attention Attention control Decision-making |
Environmental context and resources | Resources/material resources (availability and management) Environmental stressors Person × environment interaction Knowledge of task environment |
Social influences | Social support Social/group norms Organisational development Leadership Team working Group conformity Organisational climate/culture Social pressure Power/hierarchy Professional boundaries/roles Management commitment Supervision Inter-group conflict Champions Social comparisons Identity; group/social identity Organisational commitment/alienation Feedback Conflict—competing demands, conflicting roles Change management Crew resource management Negotiation Social support: personal/professional/organisational, intra/interpersonal, society/community Social/group norms: subjective, descriptive, injunctive norms Learning and modelling |
Emotion | Affect Stress Anticipated regret Fear Burn-out Cognitive overload/tiredness Threat Positive/negative affect Anxiety/depression |
Behavioural regulation | Goal/target setting Implementation intention Action planning Self-monitoring Goal priority Generating alternatives Feedback Moderators of intention-behaviour gap Project management Barriers and facilitators |
Nature of the behaviours | Routine/automatic/habit Breaking habit Direct experience/past behaviour Representation of tasks Stages of change model |
Version 2 | |
Domain (definition) | Constructs |
1. Knowledge (An awareness of the existence of something) | Knowledge (including knowledge of condition/scientific rationale) Procedural knowledge Knowledge of task environment |
2. Skills (An ability or proficiency acquired through practice) | Skills Skills development Competence Ability Interpersonal skills Practice Skill assessment |
3. Social/professional role and identity (A coherent set of behaviours and displayed personal qualities of an individual in a social or work setting) | Professional identity Professional role Social identity Identity Professional boundaries Professional confidence Group identity Leadership Organisational commitment |
4. Beliefs about capabilities (Acceptance of the truth, reality or validity about an ability, talent or facility that a person can put to constructive use) | Self-confidence Perceived competence Self-efficacy Perceived behavioural control Beliefs Self-esteem Empowerment Professional confidence |
5. Optimism (The confidence that things will happen for the best or that desired goals will be attained) | Optimism Pessimism Unrealistic optimism Identity |
6. Beliefs about Consequences (Acceptance of the truth, reality, or validity about outcomes of a behaviour in a given situation) | Beliefs Outcome expectancies Characteristics of outcome expectancies Anticipated regret Consequents |
7. Reinforcement (Increasing the probability of a response by arranging a dependent relationship, or contingency, between the response and a given stimulus) | Rewards (proximal/distal, valued/not valued, probable/improbable) Incentives Punishment Consequents Reinforcement Contingencies Sanctions |
8. Intentions (A conscious decision to perform a behaviour or a resolve to act in a certain way) | Stability of intentions Stages of change model Transtheoretical model and stages of change |
9. Goals (Mental representations of outcomes or end states that an individual wants to achieve) | Goals (distal/proximal) Goal priority Goal/target setting Goals (autonomous/controlled) Action planning Implementation intention |
10. Memory, attention and decision processes (The ability to retain information, focus selectively on aspects of the environment and choose between two or more alternatives) | Memory Attention Attention control Decision making Cognitive overload/tiredness |
11. Environmental context and resources (Any circumstance of a person’s situation or environment that discourages or encourages the development of skills and abilities, independence, social competence and adaptive behaviour) | Environmental stressors Resources/material resources Organisational culture/climate Salient events/critical incidents Person × environment interaction Barriers and facilitators |
12. Social influences (Those interpersonal processes that can cause individuals to change their thoughts, feelings, or behaviours) | Social pressure Social norms Group conformity Social comparisons Group norms Social support Power Intergroup conflict Alienation Group identity Modelling |
13. Emotion (A complex reaction pattern, involving experiential, behavioural, and physiological elements, by which the individual attempts to deal with a personally significant matter or event) | Fear Anxiety Affect Stress Depression Positive/negative affect Burn-out |
14. Behavioural regulation (Anything aimed at managing or changing objectively observed or measured actions) | Self-monitoring Breaking habit Action planning |
Use of the TDF in published implementation research
-
Identifying influences on behaviours. Exploration of barriers and facilitators to implementing specific evidence-based behaviours. Examples of interview studies include investigating facilitators and barriers to offering a family intervention to families of people with schizophrenia [24], transfusing with red blood cells [25, 26], discussing human papillomavirus (HPV) vaccination with patients [27], routinely ordering pre-operative tests [28], error-free prescribing [29], managing acute low back pain without ordering an X-ray [30], dementia diagnosis and management [31] and mild traumatic brain injury management [32]. Examples of questionnaire studies include investigating facilitators and barriers to hand hygiene [33], providing tobacco use prevention and cessation counselling among dental providers [34] and midwives engaging with pregnant women to stop smoking [35].
-
Systematic intervention design. Examples include GPs, physiotherapist and chiropractors to manage acute low back pain [36, 37]; emergency department staff management of mild traumatic brain injury [38]; hospital clinician adherence to national guidelines on the management of suspected viral encephalitis [39]; and implementation of guidelines to promote safe use of nasogastric tubes [40].
-
Process evaluations of randomised trials to better understand the effect of implementing evidence, e.g. in the Canadian CT Head Rule trials among emergency physicians [41].
Methods
-
Review current evidence for TDF
-
Identify gaps in the evidence and develop a plan to build an international collaboration among researchers and decision-makers interested in advancing the use of TDF
-
Outline an agenda for a series of studies focused on the TDF
Results
Stage | Detail | Key considerations |
---|---|---|
1. Select and specify the target behaviour/s | Use documentary analysis or empirical research to identify and specify who should do what differently, to increase the uptake of evidence-based practice | May require assessment of the feasibility of measuring the behaviour as an outcome variable |
2. Select the study design | May involve semi-structured individual interviews, focus group interviews, questionnaires, structured observations, documentary analysis or consensus processes | Design should fit the research question and will depend on the stage of investigation through exploration and development to intervention and explanation |
3. Develop study materials | Although materials from previous studies may be used as templates, materials should be adapted to be appropriate to the specified behaviour/s and context | Requires in-depth understanding of the theoretical content of each domain Requires pilot testing for comprehensibility and clinical sensibility |
4. Decide the sampling strategy | For exploratory studies, a maximum variation approach is appropriate | Key participants are those who will, or should, perform the target behaviour but other stakeholders (e.g. managers, co-workers) may also contribute a valuable perspective |
5. Collect the data | Published studies have used audio-recorded interviews (face-to-face or telephone; one-to-one or focus group) or questionnaires (paper-based or online) | Effective interviewing requires standard interviewer competencies and in-depth understanding of the theoretical content of each domain |
6. Analyse the data | The objective is to identify the domains that are most relevant to the implementation problem being addressed and to populate those domains with context-relevant and behaviourally specific content | Coding in qualitative studies requires in-depth understanding of the theoretical content of each domain |
7. Report findings | For interview studies, report presents tables that include illustrative quotations, specific beliefs identified (with frequencies, if appropriate) and classification into domains | The explanatory text relating to the table of course relates to the study objectives |
Study title |
Evaluation of a TDF-informed implementation intervention for the management of acute low back pain in general medical practice |
Rationale for changing behaviour |
Management of low back pain in general medical practice is common, but this management is not always concordant with recommended evidence-based guidelines. In particular, x-rays are overused which leads to unnecessary harm due to radiation exposure and possible detection of incidental irrelevant findings, and an intervention of known effectiveness, giving advice to stay active, is underused. |
Study design and materials |
Three phase study: 1. Qualitative methods: focus groups with general practitioners (GPs) (n = 42) using TDF to identify barriers to and facilitators of two evidence-based target behaviours related to the management of acute low back pain: one related to diagnosis, that plain film x-rays are necessary only if fracture is suspected, and one related to treatment, that of providing advice to stay active, including the avoidance of advising more than two days of bed rest. Here is an example of specifying these behaviours using the criteria: Who is performing the behaviour? What do they need to do? When do they need to do it? Where do they need to do it? If applicable, the behaviour should also be specified in terms of how often and with whom it should be done.
Behaviour 1: Manage patients without referring for plain X-ray
Who–GPs
What–Manage patients with acute low back pain without referring for plain X-ray
When–On assessment or review of patients presenting with acute, uncomplicated low back pain of less than 3 months duration and without any serious underlying pathology suspected
Where–In clinic
How often–On assessment and review
With whom–Behaviour not depended on others
Behaviour 2: Provide advice to stay active
Who–GPs
What–Provide advice to stay active
When–When managing patients with acute, uncomplicated low back pain of less than 3 months duration and without any serious underlying pathology suspected
Where–In clinic
How often–On assessment and review
With whom–Behaviour not depended on others 2. Intervention development: mapping of barriers and facilitators within TDF domains to behaviour change techniques (detail provided in French et al. [36]). The TDF was used to guide the choice of behaviour change techniques and intervention components. 3. Cluster randomised trial: evaluation of a TDF-based intervention compared to simple dissemination of the guideline (results provided in French et al. [69]). Outcomes measured included behavioural predictors (e.g. knowledge, attitudes and intentions), fear avoidance beliefs, behavioural simulation (clinical decision about vignettes) and rates of X-ray and CT-scan (medical administrative data). Forty seven practices (53 GPs) were randomised to the control and 45 practices (59 GPs) to the intervention. |
Findings and conclusions |
The TDF allowed for the systematic identification of multiple barriers and facilitators in general medical practice and subsequent mapping to behaviour change techniques. The intervention consisted of interactive workshops designed to improve the knowledge, skills, intentions and clinical decision-making of the general practitioners. The intervention had some influence on GP adherence to an evidence-based guideline for the management of lower back pain at 12 months post-intervention. Overall, the intervention led to small changes in GP intention to practice in a manner consistent with an evidence-based guideline, but it did not result in statistically significant changes in actual behaviour measured via administrative data. |
Study outputs |
Study title |
The demonstration of a theory-based approach to the design of localized patient safety interventions |
Rationale for changing behaviour |
Between 3.7 and 17.7% of patients in hospital are inadvertently harmed either by healthcare professional error or deviations from recommended practice. In this example, the TDF was used to understand behaviours related to implementing a patient safety guideline promoting safe nasogastric feeding. |
Study design and materials |
The Influences on Patient Safety Behaviours Questionnaire IPSBQ [48], a 34-item tool based on the 12-domain version of the TDF was completed by staff in three hospitals to identify influences on locally identified target behaviours relating to safe nasogastric feeding. MANOVA was used to identify highest scoring domains. |
Findings and conclusions |
Social influences, environmental context and resources, skills and emotion were identified as the most influential domains. Relevant domains were further explored in focus groups and intervention strategies generated using explicit links between theoretical domains and behaviour change techniques [7]. |
Study outputs |
Taylor et al. [72] |
Study title |
Identifying barriers to primary care type 2 diabetes management: qualitative systematic review |
Rationale for changing behaviour |
There is broad consensus and a strong evidence base to guide the care of diabetes. Despite encouraging trends in the delivery and outcomes of care for people with diabetes, there remains significant scope for improvement. Most clinical management of diabetes now occurs in primary care. Interventions to enhance the implementation of evidence-based guidelines to improve the care of people with diabetes have shown small to modest effects. To ensure that interventions address barriers to behaviour change and build on known facilitators, it is important to understand primary care clinicians’ beliefs around their day-to-day management of such patients. |
Study design and materials |
Systematic review of qualitative studies, including searches of following databases from 1980 to 2013: MEDLINE, EMBASE, CINAHL, PsycINFO and ASSIA. Qualitative studies examining diabetes management in primary care were eligible. Following screening of abstracts and full texts, data were coded to TDF domains and other themes if required. This review focused on behaviours to address clinical targets (including control of blood sugar, cholesterol and blood pressure) and processes of care (including foot examination). Findings were synthesised to identify barriers and facilitators common across or unique to clinical management goals, as well as apparent and potentially unexplored gaps in the literature. |
Findings and conclusions |
Out of 32 included studies; 17 address general diabetes care, 11 glycaemic control, three blood pressure, and one cholesterol control. Clinicians struggle to meet evolving treatment targets within limited time and resources and are frustrated with resulting compromises. They lack confidence in knowledge of guidelines and skills, notably initiating insulin and facilitating patient behaviour change. Changing professional boundaries have resulted in uncertainty about where clinical responsibility resides. Accounts are often couched in emotional terms, especially frustrations over patient adherence and anxieties about treatment intensification. |
Study outputs |
Rushforth et al. [73] |
Study title |
Understanding effects in reviews of implementation interventions using the Theoretical Domains Framework |
Rationale for changing behaviour |
There is evidence that two behaviours related to post-fracture management of patients at risk of osteoporosis are sub-optimally performed: 1) primary and secondary healthcare professionals scanning bone mineral density and 2) prescribing anti-resorptive therapy (bisphosphonate medication). This study used the TDF to identify which theoretical factors were targeted in a systematic review of interventions to improve quality of care in post-fracture investigation and their relation to observed effect sizes. |
Study design and materials |
A behavioural scientist and a clinician independently coded TDF domains in intervention and control groups in 10 interventions identified in a systematic review. For example, part of an intervention describing an ‘algorithm for diagnosis and treatment of osteoporosis’ was coded in the domain memory, attention and decision processes. Pearson’s correlations were used to explore the relationship between intervention effect size and total number of domains identified in reviews. |
Findings and conclusions |
The five domains coded most frequently (in order of frequency highest to lowest) were: 1. Memory, attention and decision processes 2. Knowledge 3. Environmental context and resources 4. Social influences 5. Beliefs about consequences Correlational analysis identified a statistically significant inverse relationship between both the domain count and frequency with the observed effect size in interventions for scanning bone mineral density, i.e. interventions with a small number of domains coded infrequently tended to have larger effect sizes than interventions with a greater number of domains coded more frequently. This relationship was not observed for interventions to improve bisphosphonate prescribing. |
Study outputs |
Little et al. [74] |
Study title |
A study of the perceived risks, benefits and barriers to the use of selective decontamination of the digestive tract (SDD) in adult critical care units |
Rationale for changing behaviour |
Critically ill patients who require management in an Intensive Care Unit (ICU) are particularly susceptible to hospital acquired infections which are associated with high morbidity and mortality. SDD may reduce these infections and improve mortality but has not been widely adopted into practice. Adoption of SDD would involve a set of protocolised behaviours performed by a range of healthcare professionals, so this investigation sought the views of multiple professional stakeholders. |
Study design and materials |
A four-phase study in three regions (the UK, Canada and Australia/New Zealand) of which Phase 2 was a Delphi study. Round 1 of the Delphi study involved one-to-one telephone interviews based on the TDF. Four key clinician groups (ICU physicians, ICU pharmacists, infectious disease clinicians/medical microbiologists, ICU clinical leads/nurse managers) were sampled using databases within each region. The researchers aimed for 10 from each group in each region. Purposive diversity sampling was used to identify a wide range of views, based on the following variables: • Hospital is academic-affiliated or not • Years of experience (time working in intensive care) • Size of ICU (number of beds) • Current practice (routinely perform SDD or not) Potential participants were ranked according to these variables and invited to participate in the Delphi study based on their ranking. During the interview phase, diversity on these factors was tracked using a diversity sampling table. |
Findings and conclusions |
141 participants were interviewed. Beliefs about Consequences was the most populous domain. “SDD increases antibiotic resistance”, “SDD reduces Ventilator Associated Pneumonia” and “SDD benefits the patients to whom it is delivered” were the most frequently mentioned beliefs, illustrating the problematic balance between potential harms and benefits. |
Study outputs |
Study title |
Anaesthesiologists’ and Surgeons’ Perceptions about Routine Pre-operative testing in low risk patients: application of the Theoretical Domains Framework (TDF) to identify factors that influence physicians’ decisions to order pre-operative tests. |
Rationale for changing behaviour |
Routine pre-operative tests for anaesthesia management are ordered by both anaesthesiologists and surgeons for healthy patients undergoing low-risk surgery, often without any clinical indication and the subsequent test results are rarely used. Identifying factors that influence why anaesthesiologists’ and surgeons’ order these routine tests for healthy patients undergoing low risk surgery provide more effective targets for intervention development. |
Study design and materials |
Interview study–sixteen clinicians (eleven anaesthesiologists and five surgeons) throughout Ontario were recruited. An interview guide based on the TDF was developed to identify beliefs about pre-operative testing practices. Physicians’ statements were content analysed into the relevant theoretical domains. Two researchers coded interview participants’ statements into the relevant theoretical domains. The first pilot interview was coded in tandem to develop the coding strategy and the second was used to ensure the two coders were comfortable with the strategy developed from the first. Subsequent coding of the remaining interviews was completed independently and Fleiss’s Kappa (κ) was calculated for all domains and interviews to assess whether the two researchers coded the same text into the same domain. Within each domain, the primary coder wrote a belief statement that captured the core thought of each utterance. For example, the following utterances were coded under the domains Social Influences: “… if a surgeon ordered it I am somewhat reluctant to cancel one of their tests even though I don’t feel that it’s necessary” & “Sometimes they are ordered and then (we) might be reluctant to cancel some of the tests because I am not privy to their thought process….”. These 2 utterances were from 2 different respondents but reflect the same core thought: I’m reluctant to cancel tests ordered by other physicians. Identical beliefs statements were then grouped together. Statements that centred on same theme or were polar opposites of a theme were also grouped together for the ease of further analysis. For example, the following 3 belief statements from Social Influences grouped under the theme influence of colleagues: The opinions of others do not influence my decision to order routine tests. I’m reluctant to cancel test ordered by other physicians. I order tests I feel are unnecessary because my conservative colleague may be in the operating room on the day of the surgery and want to see the routine test that I would not. Belief statements that were coded in different domains by the researchers were discussed to establish consensus. Where single domain allocation agreement could not be reached, researchers agreed that the statement could be placed in both domains. |
Findings and conclusions |
Seven domains were identified as likely relevant to changing clinicians’ behaviour about pre-operative test ordering for anaesthesia management (Social/professional role and identity, Beliefs about capabilities and Social influences, Environmental context and resources, Beliefs about consequences, Behavioural regulation, Nature of the behaviour). Key beliefs identified within these domains included: conflicting comments about who was responsible for the test-ordering, inability to cancel tests ordered by fellow physicians, and the problem with tests being completed before anaesthesiologists see patients. Anaesthesiologists often ordered tests based on who may be the attending anaesthesiologist on the day of surgery while surgeons ordered tests they thought anaesthesiologists might need. There was also a range of comments about the consequences associated with reducing testing, from negative (delay or cancel patients’ surgeries), to indifference (little or no change in patient outcomes), to positive (save money, avoid unnecessary investigations). |
Study outputs |
Patey et al. [28] |
Study title |
A cross-country comparison of intensive care physicians’ beliefs about their transfusion behaviour: A qualitative study using the theoretical domains framework. |
Rationale for changing behaviour |
Transfusion of blood, a scarce and costly resource, is used for treating a variety of medical conditions. There is a wide variation in blood transfusion behaviour across different medical disciplines including intensive care physicians. A restrictive transfusion is, at least, equivalent and possibly superior to a more liberal transfusion. The aim of the study was to elicit beliefs about specified behaviour within each theoretical domain and role of the domain in influencing the behaviour in intensive care units across Canada. |
Study design and materials |
Ten intensive care physicians throughout Canada were interviewed. Physicians’ responses were coded into theoretical domains, and specific beliefs were generated for each response. Theoretical domains relevant to behaviour change were identified if they included belief statements that might be potential barriers for changing transfusion behaviour and fulfilled the following criteria: (1) relatively high frequency of specific beliefs, (2) presence of conflicting beliefs, and (3) evidence of strong beliefs that may impact on the behaviour. All three criteria were considered concurrently to judge relevance of the domains. Beliefs within the domains were analysed for psychological constructs and were subsequently used to select psychological theories using the methodology proposed by Francis et al. [25]. |
Findings and conclusions |
Seven theoretical domains populated by 31 specific beliefs were identified as relevant to the target behaviour using all criteria. The relevant theoretical domains were Knowledge, Social/professional role and identity, Beliefs about capabilities, Beliefs about consequences, Motivation and goals, Social influences and Behavioural regulation. For example, Knowledge domain was identified as potentially relevant because majority participants reported the belief that there is not enough evidence to support watching and waiting in all patient populations. Motivation and goals was identified as a key domain because conflicting specific beliefs were elicited (e.g. Watching and waiting conflicts with other goals in opposition to Watching and waiting is compatible with other goals). When the belief that ‘emotion does not affect my decision to transfuse’ was consistently reported, it was concluded that the Emotion domain was not relevant to the transfusion behaviour. For greater detail please see the published article. |
Study outputs |
Islam et al. [26] |
Time estimates for conducting research using TDF
-
Steps 1–3: Selecting and specifying the target behaviour, selecting study design, and deciding the sampling strategy may take days or weeks. In relation to identification of the target behaviour, conducting interviews and follow-up work will have cost implications so there needs to be good evidence that changing the behaviour in question will produce benefits and reduce harms.
-
Step 4: Developing study materials may take weeks to months to produce, pilot and finalise interview schedules and topic guides for focus groups.
-
Step 5: Collecting data is likely to take months to complete. In some cases, it may take 1 month to complete interviews but it can easily take several times longer depending on the numbers required and difficulties with recruitment.
-
Step 6: Analysing the data may take months and will depend on the amount of data, number of staff coding the data and number of disagreements in coding the data.
Report findings
Discussion
Potential applications of the TDF
Linking TDF to other theoretical models
Study title |
Factors Influencing Variation in Physician Adenoma Detection Rates: a Theory-Based Approach for Performance Improvement. |
Rationale for changing behaviour |
Interventions to improve physician adenoma detection rates (ADRs) for colonoscopy have generally not been successful. There is limited understanding of which factors influence variation which might be appropriate targets for intervention. |
Study design and materials |
Three focus groups of gastroenterologists and three of endoscopy nurses were conducted at medical centres in Northern California. As participants were available for a limited time (45–60 minutes), an adaptive interviewing method was used. First, participants were asked questions covering the three components of the COM-B model (capability, opportunity and motivation) to identify factors relevant in explaining ADR variation. Then for each relevant COM-B component, participants were asked questions covering the related domains of the TDF. For example, to investigate participants’ capabilities to perform a behaviour, they were asked “would you be more/less likely to do ‘X’ if you had greater physical and/or psychological ability?” If they responded positively, the researcher asked further questions structured by TDF domains representing capability, i.e. knowledge; physical skills; memory, attention and decision processes and behavioural regulation. |
Findings and conclusions |
This adaptive interviewing method optimised the time available with higher level COM-B questions acting as a filter to potentially relevant TDF domains. |
Study outputs |
Atkins et al. [81] |