Article Text
PDF
Abstract
Objectives The Conversation, Understand, Replace, Experts and evidence-based treatment (CURE) project implemented an evidence-based intervention that offers a combination of pharmacotherapy and behavioural support to tobacco-dependent inpatients. Understanding key characteristics of CURE’s implementation strategy, and identifying areas for improvement, is important to support the roll-out of nationwide tobacco dependence services. This study aimed to (1) specify key characteristics of CURE’s exiting implementation strategy and (2) develop theoretical-informed and stakeholder-informed recommendations to optimise wider roll-out.
Design and methods Data were collected via document review and secondary analysis of interviews with 10 healthcare professionals of a UK hospital. Intervention content was specified through behaviour change techniques (BCTs) and intervention functions within the Behaviour Change Wheel. A logic model was developed to specify CURE’s implementation strategy and its mechanisms of impact. We explored the extent to which BCTs and intervention functions addressed the key theoretical domains influencing implementation using prespecified matrices. The development of recommendations was conducted over a two-round Delphi exercise.
Results We identified six key theoretical domains of influences: ‘environmental context and resources’, ‘goals’, ‘social professional role and identity’, ‘social influences’, ‘reinforcement’ and ‘skills’. The behavioural analysis identified 26 BCTs, 4 intervention functions and 4 policy categories present within the implementation strategy. The implementation strategy included half the relevant intervention functions and BCTs to target theoretical domains influencing CURE implementation, with many BCTs focusing on shaping knowledge. Recommendations to optimise content were developed following stakeholder engagement.
Conclusions CURE offers a strong foundation from which a tobacco dependence treatment model can be developed in England. The exiting strategy could be strengthened via the inclusion of more theoretically congruent BCTs, particularly relating to ‘environmental context and resources’. The recommendations provide routes to optimisation that are both theoretically grounded and stakeholder informed. Future research should assess the feasibility/acceptability of these recommendations in the wider secondary-care context.
- preventive medicine
- public health
- qualitative research
- respiratory medicine (see thoracic medicine)
Data availability statement
Data are available upon reasonable request. The data sets used and/or analysed during the current study are available from the corresponding author on reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
Strengths and limitations of this study
This study is the first to qualitatively explore behavioural factors underpinning the implementation of the Conversation, Understand, Replace, Experts and evidence-based treatment (CURE) project.
The behavioural analysis, and subsequent stakeholder involvement, has resulted in tailored, practical recommendations for optimisation of future tobacco dependence services, which facilitate efficient translation of findings into policy and practice.
Due to its early phase of roll-out, our recommendations have been developed from implementation within a single UK hospital implementing CURE, therefore generalisability of findings to other contexts may be limited.
Feedback was not gathered from patients or members of the public, therefore the barriers and facilitators of implementation and the stakeholder-informed recommendations are limited to the views of those commissioning, delivering and implementing CURE.
Introduction
The government NHS Long Term Plan1 has outlined a commitment to offer National Health Service (NHS)-funded tobacco treatment services to all those admitted to hospital by 2023/2024. However, the most recent National Smoking Cessation Audit Report from the British Thoracic Society2 suggests that adherence to national smoking cessation standards remain poor. For example, despite the expected standard being 100%, only 77% of inpatients had their smoking status recorded. Of those who smoked, just 44% were asked if they would like to quit smoking; and of those who were referred for smoking cessation support, just 16% were referred to hospital-based services (with a further 8% referred to community-based services). In addition to this, only 31% of the smokers were offered nicotine replacement therapy (NRT). As a result, the report set national improvement objectives to support and offer NRT to all inpatient smokers, and to provide further support and training to hospital staff to ensure that they are able to implement tobacco dependence treatment into their everyday practice.
Hospitalisation provides a unique opportunity to identify and engage smokers, initiate cessation treatments and facilitate appropriate follow-up and support for them.3 4 Intensive smoking cessation interventions that begin in hospital and include pharmacotherapy, counselling and post-discharge support for ≥1 month, increase the likelihood of smoking abstinence (risk ratio 1.37, 95% CI 1.27 to 1.48; 25 studies) compared with hospital-only interventions with no follow-up.4
The Ottawa Model for Smoking Cessation (OMSC), initially implemented in Canada, aims to increase the rate at which smoking cessation support is offered to all smokers within secondary care (ie, hospital settings).5 6 The OMSC provides a systematic approach to screening all inpatients for smoking status, with those who smoke offered a combination of pharmacotherapy and behavioural support. Patients are then attached to ongoing community stop-smoking support post-discharge.7 The OMSC model was found to have positive outcomes in increased smoking abstinence at 6 months, reduced all-cause re-admissions at 30 days and 1 year and reduced mortality at 1 year when compared with a control group receiving usual care.7
The positive outcomes observed in Canada led to the development of the Conversation, Understand, Replace, Experts and evidence-based treatments (CURE) and has recently been piloted within an NHS trust in the North West of England.8 Importantly, CURE aims to increase awareness about the medicalisation of tobacco dependence and encourage clinicians in offering smoking cessation care to all inpatient smokers. Similar to the OMSC, the CURE project aims to improve smoking outcomes by providing combination of pharmacotherapy (eg, NRT, varenicline) and behavioural support to patients, as well as post-discharge care at 2, 4 and 12 weeks. The CURE implementation intervention includes various strategies designed to change behaviours at organisational, practitioner or patient levels and to enhance the adoption of a clinical innovation.9 Examples of implementation strategies include outreach activities, in-house training, audit and feedback and computer prompts.
Evaluation of the CURE pilot (October 2018 to March 2019) showed that 92% of all adult admissions (total admissions: 14 690) were screened for smoking status10 with a cost per quit of £475.11 More importantly, the evaluation demonstrated a positive patient impact; out of 2293 patients identified as current smokers, 96% were provided with brief advice, 61% accepted and completed a specialist behavioural support, 66% were prescribed pharmacotherapy (eg, NRT, varenicline) to support quit attempts and 22% were abstinent at 3 months post-discharge.10 These findings suggest that the model may be useful in assisting clinicians’ behaviour change when compared with national audit data. It would therefore be valuable to determine how the CURE project was delivered in practice. This knowledge would support recommendations for a national specification model, based on the OMSC and CURE, for further testing and piloting.1
To maximise the potential benefits of CURE, there is a need to understand the implementation process of this evidence-based smoking cessation intervention in routine secondary care. Several theoretical approaches (ie, theories, models, frameworks) can be used to provide a better understanding and explanation of how and why implementation succeeds or fails.12 13 For instance, the Theoretical Domains Framework (TDF) represents an approach to understand what determinants are hypothesised to influence implementation outcomes, (eg, healthcare practitioners’ adoption of an evidence-based patient intervention).13 14 The TDF summarises 14 broad domains relevant to changing behaviour, ‘knowledge’, ‘beliefs about consequences’, ‘beliefs about capabilities’, ‘skills’, ‘environmental context and resources’, ‘social influences’, ‘memory, attention and decision processes’, ‘behavioural regulation’, ‘emotion’, ‘social or professional role/identity’, ‘optimism’, ‘intentions’, ‘goals’ and ‘reinforcement’.15 16
Another theoretical approach to explain the causal mechanisms of implementation is the COM-B (Capability, Opportunity, Motivation and Behaviour) model, which suggests behaviour is a function of physical and psychological capability, physical and social opportunity and automatic and reflective motivation. The COM-B model sits at the hub of the Behaviour Change Wheel (BCW) (see figure 1),14 17 a well-established guide, applied to health services research, to provide a systematic approach to identifying intervention content and specifying mechanisms of action (ie, how interventions elicit behaviour change).14 The wheel comprises three main ‘layers’: (1) sources of behaviour (ie, the COM-B model), (2) nine intervention functions (ie, means by which behaviour can be changed) and (3) policy categories (ie, that may support delivery of intervention functions) (p.17).14
Visual representation of the Behaviour Change Wheel.14 TDF, Theoretical Domains Framework.
When aiming to understand how behaviour may be changed and/or may specify implementation content, the intervention functions within the BCW can be linked to specific behaviour change techniques (BCTs), which are defined as ‘an active component of an intervention designed to change behaviour’. BCTs have been associated with many types of behaviour which have been brought together to form an international BCT Taxonomy V.1 with 93 BCTs.18
Theoretical approaches such as BCW, the COM-B model (figure 1), the TDF and the BCT Taxonomy (BCTTv1), may be applied in conjunction with one another to understand the implementation process, identify implementation strategy content and to explore barriers to and facilitators of behaviour. Prior research has successfully integrated these theoretical approaches to explore determinants influencing the implementation process of evidence-based practice in healthcare.16 19
When planning implementation, developing a logic model of links between implementation strategies, mechanisms and outcomes is crucial.20 The BCW facilitates the specification of outcomes, determinants, change objectives and intervention, and it thereby enables intervention developers to map specific BCTs to behavioural determinants.17
Informed by the BCW,14 the present study aimed to describe the core elements of the CURE implementation strategy in the pilot site, particularly the activities directed at promoting behaviour change in healthcare practitioners and wider organisational implementation strategies (organisational/professional level).
The specific objectives of this study were to:
Describe the content of CURE’s implementation strategy, using BCW functions, policy categories and the BCTTv1.21
Characterise the intervention in a logic model to clarify causal assumptions and mechanism of impact using the Medical Research Council (MRC) guidance.21
Explore to what extent the barriers and facilitators of CURE implementation are addressed by existing implementation strategy components.
Develop recommendations to optimise the future implementation of CURE.
This work is a first step in designing a successful theoretical-informed implementation strategy for wider, national roll-out. This work was conducted alongside a TDF-based, qualitative study which explored the barriers and facilitators of CURE implementation and delivery, from the perspective of healthcare professionals engaged in the project pilot.22
Methods
We undertook a systematic, theoretically-guided approach to specify the content and possible mechanisms of action and impact of the implementation strategy of CURE. This process has previously been coined as ‘strategic behavioural analysis’.19 We have employed the use of the StaRI (Standards for Reporting Implementation Studies) as our reporting standard.23
Setting and participants
The pilot site is a major acute teaching hospital with approximately 900 beds and 27 500 inpatient admissions per year (excluding maternity, paediatrics, Accident & Emergency (A&E) admissions and Intensive Care Unit), providing both district general hospital services and specialist tertiary services. Tertiary services include cardiology, cardiothoracic surgery, heart and lung transplantation, respiratory conditions, burns and plastics, cancer and breast care services. The smoking prevalence included in the pilot site was modelled based on 20% of inpatient admissions (approximately 5500 smokers per year).
At admission, the admitting clinicians (doctor or nurse) were responsible for recording smoking status, assessing level of addiction and offering initial rapid treatment. A CURE specialist team would then perform a visit, review all smokers admitted (opt-out service) and complete specialist assessment, update treatment plan and plan for discharge (eg, refer to community service). For the pilot study, the implementation plan modelled the need for five specialist CURE nurses to deliver the specialist assessment, treatment planning and follow-up for all smokers admitted as inpatients.
Patient and public involvement
As this study focused on healthcare professionals’ behaviour change, no patients or members of the public were involved.
Procedure and sources of data
To collect data on the implementation strategy content, we used two different methods:
Document analysis. Researchers read and re-read training materials (ie, training manual, training poster, teaching slides, level 1 and 2 eLearning modules, Steering Group Terms of Reference) and the CURE project webpage (available from https://thecureproject.co.uk/) describing implementation strategy content, including the training materials, practice tools, promotional/educational materials and smoke-free policy. We (AMR, AH and AW; health psychology specialists) reviewed and appraised documentation by systematically mapping information against the Template for Intervention Description and Replication (TIDieR)24 and the BCW components, including BCTs, intervention functions and policy categories.14 This information was also used to develop an initial logic model.
Semi-structured interviews. We conducted secondary analysis of semi-structured interview data with 10 purposively sampled healthcare professionals, who were involved in the implementation and delivery of the CURE evidence-based intervention (reported in full elsewhere; 22). Participants spanned core CURE management (n=2) and specialist nursing staff (n=3), pharmacy (n=1), primary care (n=1) and public health (n=3). Interview topic guides were informed by TDF domains and asked participants to discuss barriers and facilitators to implementing the CURE project pilot and detail implementation strategy content (ie, describing what was delivered, with what aim, how much, to whom and by whom). All interviews were digitally recorded, transcribed verbatim and analysed using the Framework Method.25 Data from interviews were also used to revise the logic model.
Data analysis
Step 1—implementation strategy content analysis
Using the TIDieR framework,24 we created a broad outline of the implementation strategy that included the content delivered, to whom and by whom, why, by what mode of delivery, how often, where, when and how much, tailoring, modifications and fidelity. Data from all data sources were used. Data collected from both the document analysis and interviews were coded for implementation strategy content (AMR, AH and AW) using existing coding frameworks provided by the BCW guide14; appendix 4 (p.259 of the guide) for BCTs, table 2.1, 14 (p.111 of the guide) to code intervention functions and table 2.7, 14 (p.135 of the guide) to code policy categories. Any discrepancies in coding were resolved via consensus discussion.
Step 2—mechanisms of impact (logic model)
Following the guidance on developing logic models in process evaluations of complex interventions, issued by the MRC,21 we developed a logic model by reviewing the CURE documentation and service specification (https://thecureproject.co.uk/), current evidence,7 8 26 and theoretical understandings of both the evidence-based intervention and the implementation strategy as suggested in the TiDIER guidelines. Public Health England liaised with the CURE project team (via email) who provided additional documentation (pathway mapping workshop slides, early evaluation options, inpatient numbers and time commitments for specialist nurses, communications plan, Tobacco Addiction Service data) to further inform the logic model. An initial logic model was reviewed and updated based on findings from the qualitative interviews and behavioural analysis demonstrating the intended mechanisms of impact (initial model) versus actual mechanisms of impact, that is, what was delivered in practice (revised model).
Step 3—identifying opportunities for optimisation
In line with previous research,19 27 the following mapping exercise was conducted in order to explore the extent to which barriers and facilitators of CURE implementation22 were addressed by existing implementation strategy components, and to identify any missed opportunities for further design:
A concurrent qualitative study22 reported eight key TDF domains that influenced CURE implementation (see online supplemental file 1 for a summary of these findings). To identify key domains influencing the implementation of CURE, we ranked these previously reported TDF/COM-B domains using established criteria: frequency (number of transcripts in which a domain occurred), elaboration (number of themes within a domain) and evidence of conflicting statements within domains (eg, if some participants report lack of specific skills whereas others report having the relevant skills).28–30 All of these factors were considered concurrently in establishing domain relevance. This process was facilitated through consensus discussion between the two researchers (AMR and AW) and supported by a third researcher to resolve any discrepancies (AH).
The outputs of the key domains and content analysis stages were combined by mapping the identified influences to the identified BCT and intervention functions of the CURE implementation strategy. This was achieved by combining two available matrices that map the TDF to the BCTTv131 32 and the Theory and Techniques Tool (https://theoryandtechniquetool.humanbehaviourchange.org/) as was developed for previous research.19 This analysis investigated the level of theoretical congruence between implementation strategy components of CURE and the qualitative data on barriers and facilitators influencing its implementation.
The level of theoretical congruence between influences on behaviour (TDF domains) and implementation strategy content to change behaviour (BCTs) was achieved by analysing the extent to which the BCTs identified in the CURE implementation strategy targeted the key TDF domains (identified in the qualitative data). Each BCT identified was coded as either low congruence (did not target any key domain), medium congruence (targeted at least one key domain) or high congruence (targeted two or more key domains).19
The mapping exercise was repeated for intervention functions and policy categories, by consulting the matrices mapping BCW against COM-B/TDF14 to identify the extent to which functions (matrix on p. 116) and policy categories (matrix on p. 138) in the CURE implementation strategy targeted key factors influencing the implementation process, and what additional intervention functions and policies may address barriers/facilitators within the key domains. The following definitions were applied:
Opportunity seized—instances where a theoretically congruent intervention function/policy category (according to the matrices) was identified in the existing CURE implementation strategy at least once.
Missed opportunity—instances where the theoretically congruent intervention function/policy category was not identified in the existing CURE implementation strategy.
Supplemental material
Step 4—development of recommendations to support future implementation
Following steps 1–3, the research team used the findings from the qualitative interviews and strategic behavioural analysis to draft a list of practical recommendations to strengthen implementation strategy content (ie, content likely to encourage healthcare professional behaviour change and support implementation of a secondary care-based tobacco dependence treatment model). These recommendations included example strategies to deliver BCTs relevant to the key TDF domains. To enhance the suitability and acceptability of these recommendations, a Delphi study was conducted by collecting data from a panel of six experts until consensus was reached.33 Experts included the CURE management team, PHE Programme Managers (eg, Tobacco Control and NHS Long Plan) and NHS England representatives. The six experts independently rated whether each recommendation was affordable, practical, effective, acceptable, safe and equitable (the APEASE criteria),14 on a dichotomous scale of yes,1 no/uncertain (0) for each criteria. This gave a total possible score of 36 for each recommendation. These ratings were then used to structure and encourage discussion surrounding uncertainties and potential modifications during a collaborative, stakeholder workshop. A total of 11 stakeholders participated in the stakeholder workshop. Participants included two members of the research team (one workshop facilitator and one scribe), two members of the CURE management team, four PHE Programme Managers (eg, Tobacco Control and NHS Long Plan), one representative from NHS England and two consultants. Workshop feedback was incorporated into a refined recommendations table, which was then circulated via email for further stakeholder comment and review. This process resulted in the final list of recommendations.
Results
Step 1—implementation strategy content
Table 1 summarises the content of the implementation strategy, using the TIDieR framework. The following broad components of CURE implementation strategy were identified: staff training, practice tools, reminder systems, educational outreach visits, audit and feedback, primary care incentives, use of a steering group, branding materials, clinician implementation team meetings to promote reflective discussion, provision of local technical assistance (eg, admin support), promotion of network weaving (eg, information sharing), physical environment changes (eg, consultation facilities) and a triage system.
TIDieR table for the CURE project implementation strategy in the pilot site
Through content coding we identified 26 BCTs (ie, ‘active components’), 5 intervention functions and 4 policy categories. Further details of these activities, BCTs, intervention functions and policy categories can be found in table 2.
BCTs, intervention functions and policy categories identified in CURE
Step 2—mechanisms of impact (logic model)
The initial model is presented in figure 2. The original logic model, based on the CURE implementation strategy, shows all patients who are admitted to hospital should be asked whether they smoke, and their response should be recorded in the hospitals’ electronic patients records. All smokers should be offered immediate NRT and specialist support through motivational interviewing and behavioural change support as well as access to additional evidence-based pharmacotherapy treatments for tobacco addiction. All smokers should be offered further appointments with a specialist team after discharge from hospital to continue their support.
CURE stop smoking project: Initial logic model. CURE, Conversation, Understand, Replace, Expert and evidence-based treatment; IT, information technology; NRT, nicotine replacement therapy; PREOP, preoperative.
The logic model was reviewed and updated iteratively based on findings from the qualitative interviews and behavioural analysis. The final model is presented in figure 3. The final logic model contains further facilitators identified as important by key stakeholders (eg, funding, tobacco policy, nurse champion) as well as clarification of the meaning of an adequately resourced and staffed implementation strategy (eg, office space, clerical support, phone/computer access). Other local stakeholders essential to the smooth implementation and delivery of CURE were also added to the revised model (eg, Clinical Commissioning Group; Local Medical Committee (LMC); local General practitioner) as well as barriers to successful implementation and delivery (eg, staff turnover, staff confidence, paperwork). While a structured protocol and treatment pathway was an important facilitator, the final model includes more detail regarding the potential variety of patient journeys and the role of hospital pharmacy. The importance of patient choice was added to the final model, because it was highlighted as important to both choices of NRT and of the discharge pathways. However, there were many challenges to implementing many of the pathways as intended. This tension between primary and secondary care was highlighted in the final model.
CURE stop smoking model: Final logic model following stakeholder consultations and behavioural analysis. CURE, Conversation, Understand, Replace, Expert and evidence-based treatment; IT, information technology; NRT, nicotine replacement therapy; PREOP, preoperative.
Step 3—identifying opportunities for optimisation
Previously identified TDF/COM-B domains influencing implementation are summarised in online supplemental file 1. Considering the frequency, elaboration of the domains and evidence of conflict, the following six domains were considered the key domains of influence relating to the implementation strategy; (i) Environmental Context and Resources (Physical Opportunity; eg, integration with the wider healthcare context, staffing resources, hospital delivery environment, availability of CURE-related knowledge and training, CURE branding and flexibility of the service specification), (ii) Goals (Reflective Motivation; eg, promoting CURE, adhering to a CURE service specification, identifying and evaluating outcomes), (iii) Social Influences (Social Opportunity; eg, peer support, CURE champions, organisational culture change), (iv) Reinforcement (Automatic Motivation; eg, reflection on intrinsic rewards related to CURE involvement and delivery), (v) Social Professional Role and Identity (Reflective Motivation; eg, commitment to patient choice, acceptance of responsibility for delivering tobacco dependence treatment) and (vi) Skills (Psychological Capability and Physical Capability; eg, previous experience and skills supporting smoking cessation and using hospital-based information technology (IT) systems). These domains acted as both barriers and facilitators to implementation. Based on the criteria, we suggest these six key domains are prioritised for change (see table 3).
Prioritisation of TDF domains for the implementation of the CURE model by frequency, thematic elaboration and evidence of conflicting beliefs
Of the 26 BCTs identified in the current implementation strategy content, 6 had high theoretical congruence with the key domains identified above, 9 had medium congruence and 11 BCTs had low theoretical congruence (see table 4). The BCTs observed to have high theoretical congruence were (i) social support (practical), (ii) social support (emotional), (iii) social support (unspecified), (iv) reward (outcome), (v) restructuring the social environment and (vi) demonstration of the behaviour. These BCTs were paired with domains rated as important in influencing CURE implementation. For instance, the domain Social influences (eg, peer support, visibility of CURE champions) was appropriately targeted via the BCT Social support (practical), delivered through the implementation strategy component educational outreach visits (whereby nurse leads, clinical leads and/or CURE nurses visit colleagues, providing information and advice to support their ability to engage with CURE).
Theoretical congruence between the BCTs identified in CURE implementation strategy content and the key TDF domains influencing implementation of CURE within the pilot site
Table 5 shows whether intervention functions identified in the CURE implementation strategy appropriately targeted the six most important TDF/COM-B components. The potential missed opportunities (eg, as highlighted by the analysis) were related to the intervention functions Coercion and Restriction, which were not identified in the CURE implementation strategy. The Coercion intervention function may have been useful in targeting the domains linked to Reflective Motivation addressing themes under the TDF domain ‘Goals’ such as Managing competing goals and priorities and Promoting CURE. Nevertheless, other intervention functions were used to target this component: Education, Incentivisation and Persuasion. The Restriction intervention function may have been useful in targeting Environmental Context and Resources (Physical Opportunity) and Social Influences (Social Opportunity). Other intervention functions were used to target these TDF/COM-B components: Enablement, Environmental restructuring, Training, and Modelling.
Seized and missed opportunities: intervention functions linked with CURE
Table 6 shows whether intervention functions identified in the CURE implementation strategy were delivered through policy categories suggested by the BCW intervention function/policy category matrix. All intervention functions were delivered through at least one policy category suggested by the matrix.
Seized and missed opportunities: policy categories linked with CURE
There were missed opportunities to deliver functions identified in implementation strategy through the policy category of fiscal measures, regulation and legislation. This was particularly important for the Training (one out of four opportunities were ‘seized’) and Environmental restructuring (two out of five opportunities were ‘seized’) intervention functions, as they could have been better supported by including these policy categories.
Step 4—development of recommendations to support future implementation
Following stakeholder involvement, the final list includes 29 recommendations. Table 7 presents the final overview of recommendations, with a brief indication of stakeholder APEASE evaluations.
Recommendations to support the implementation of a nationwide, secondary care-based tobacco dependence treatment model, based on the CURE project
Initially, 26 recommendations were developed to address the themes identified within the six most important TDF domains. Recommendation ratings from the Delphi survey ranged from 3 to 36 (maximum score) with a median of 28.5 (IQR, 25.25–31.0). Survey responses are available in online supplemental file 2. These ratings were used to structure discussion within the subsequent stakeholder workshop. The workshop focused predominately on recommendations which had greatest levels of uncertainty, further contextualised these recommendations considering the existing healthcare system and specified the feasibility of implementing recommendations in practice. This included the removal of a recommendation related to financial incentives for GPs (ie, Provide financial incentive on performance (eg, when prescribing NRT) for primary care staff supporting service outpatients in the community). This was the lowest rated recommendation within the Delphi survey, with further stakeholder discussion suggesting financial incentives were not deemed acceptable nor considered effective within the pilot phase. Another recommendation relating to the delivery environment (ie, Ensure adequate facilities are available to support delivery, including physical spaces for one-to-one sessions, hospital accessibility for patients (ie, through parking, public transport) and vaping facilities) was thought to cover a lot of separate components and thus was separated into three recommendations covering the need to provide (1) adequate office space for delivery staff, (2) physical space to deliver one-to-one support to patients and (3) on-site vaping facilities. Access to IT equipment (eg, laptops), was also added as a recommendation in light of increased need to self-isolate due to the COVID-19 pandemic. A highly rated recommendation relating to deliverers’ skill development (ie, Provide additional training on how to use tools associated with intervention delivery, so staff practice and observe use of these tools to facilitate day to day delivery) was expanded to support deliverers capacity to provide behavioural support to patients. As such, an additional recommendation (to allow deliverers to shadow experienced staff members) was added, as this was identified as a facilitator of delivery during the pilot phase.
Supplemental material
Discussion
Summary of findings
This study aimed to specify the content of CURE’s implementation strategy and to develop theory-based recommendations to optimise future implementation of secondary-care/hospital-based tobacco dependence services. The existing implementation strategy incorporated half the potentially relevant content to target key identified barriers and facilitators for the CURE project. However, there were missed opportunities to further facilitate implementation as a large proportion of the BCTs within the current implementation strategy focused on the TDF domain ‘knowledge’. These findings highlight that some of the implementation strategy features were primarily educational, though many of the barriers related to the social and environmental context. More theoretically congruent BCTs should be included in the implementation strategy, particularly for the TDF domains ‘environmental context and resources,’ ‘social professional role and identity’ and ‘social influences’. The recommendations presented within table 7 highlight potentially feasible ways in which these BCTs could be operationalised.
The study used a systematic, theoretically-guided approach to specify the content and possible mechanisms of action of an implementation strategy using behavioural science methodology and triangulation from different data sources (ie, semi-structured interviews, document analysis, Delphi survey, stakeholder engagement). We have also illustrated how theory can be used to optimise the implementation strategy of the CURE project. From interviews with healthcare professionals, six themes were identified as influences for the implementation of CURE.22 These were used to identify gaps in the existing implementation strategy and informed recommendations for refinement. The implementation strategy consisted of 26 BCTs (ie, ‘active components’), 7 intervention functions and 4 policy categories that could stimulate behaviour change through several mechanisms of action, especially ‘beliefs about consequences’ (Reflective Motivation) and ‘knowledge’ (Psychological Capability). Similarly, previous systematic reviews have shown that educational strategies were the most commonly used strategies in multi-strategy interventions.34 35 Current evidence suggests that organisational-level interventions in the healthcare context can influence clinical outcomes and efficiency.36 When used as part of multi-strategy interventions, group education and organisational strategies (eg, creation of an implementation team) corresponded with positive significant changes in outcomes.34 Incorporating theory12 in the design of implementation strategies would enhance the field’s understanding of the causal mechanisms by which the strategies lead, or do not lead, to changes in outcomes at all levels.
The logic model specifies the theory of change related to mechanisms, assumptions and outcomes of the CURE model. The initial version of the model (as presented in figure 2. CURE stop smoking project: Initial logic model) presents the intended process of change, as informed by the document review. The final iteration of the model (as presented in figure 3) demonstrates a more accurate overview of what ultimately was delivered in the programme, and documents the actual process of change, as informed by document review, stakeholder views and behavioural analysis.
Several challenges to adoption and implementation of the Ottawa model have been identified previously (Reid et al, 2010). Likewise, these challenges typically included staff regarding smoking as a ‘lifestyle choice’ and a lack of support from key opinion leaders and clinical managers. Leadership and performance feedback from managers, training about tobacco-dependence treatment and smoke-free hospital policies were the key recommendations to improve adoption and implementation (Reid et al, 2010). This evidence base has been used to underpin the delivery of smoking cessation in secondary care settings, and to inform future implementation strategies.37
Other studies have successfully integrated similar theoretical approaches (ie, BCW, TDF) and methodologies (eg, qualitative interviews, Delphi, stakeholder involvement) to characterise the content and theoretical mechanisms of action of an existing implementation strategy, and to optimise an existing implementation strategy.38 39 The findings from this strategic behavioural analysis are similar to those of other studies, particularly that only a small percentage of BCTs used in interventions (21%–37.5%) are theoretically relevant for targeting identified barriers to deliver or implement behaviour change interventions.18 29 Likewise, missed opportunities in the implementation strategy content are similar across other behavioural analyses that highlighted that most focus on shaping knowledge rather than addressing motivational, social and environmental influences.18 29
This study provides relevant evidence to further guide the implementation process and selection of strategies; ensuring that enough attention is paid to planning implementation; and a flexible approach that allows response to emerging barriers, particularly at the organisational level. According to Li et al40 organisational contextual features (eg, organisational culture; leadership; networks and communication; resources; evaluation, monitoring and feedback; and champions) were most commonly reported to influence implementation outcomes across a wide range of healthcare settings.
Strengths and limitations
This study is the first to qualitatively explore behavioural factors underpinning the implementation of the CURE project. Considering barriers and facilitators to implementation through the lens of the TDF allows for the identification of both internal and external factors which are known to influence behaviour change and evidence-based intervention implementation. The behavioural analysis links these barriers and facilitators to specific components underpinning the CURE implementation strategy. This therefore provides novel insight into key factors which can facilitate implementation of such an intervention in a hospital setting. The NHS long-term plan aims to roll-out adaptations of the CURE and Ottawa models across acute, maternity and mental health settings.1 As such, this study is further informing and supporting implementation of NHS-funded tobacco dependence services in England.41 Given the time and financial constraints of this study (conducted during the early stages of the COVID-19 pandemic), and the focus on facilitating healthcare professionals’ implementation behaviour, stakeholder consultation was limited to healthcare professionals. As such, patients or the public were not involved in the development of this research. The inclusion of patient perspectives should therefore be prioritised in future work.
Due to its early phase of roll-out, our recommendations were developed from data relating to a single UK hospital implementing CURE. As such, generalisability of findings to other contexts may be limited. From these findings, relevant decision-makers can make a strategic, informed decision using evidence-based recommendations to optimise the implementation and delivery of future NHS-funded tobacco dependence treatment and target mechanisms of healthcare professional’s behaviour change. This approach also provides further insight into potentially overlooked, yet relevant, intervention functions (ie, missed opportunities) which may be considered by decision-makers to optimise the implementation of secondary care-based tobacco dependence services. Overall, the systematic approach taken throughout the present research, and use of established theoretical frameworks, results in evidence which, importantly, facilitates efficient translation to policy and practice.14
Implications for practitioners, policymakers and future research
Based on the appraisal of the CURE implementation strategy content, the current package shows good practice for implementation including relevant BCTs, intervention functions and policy categories. However, the additional recommendations provided may optimise and inform future implementation. This is a set of practical recommendations co-developed with stakeholders and informed by robust behaviour change theoretical approaches.
The BCTs currently in use are linked to multiple intervention functions, including the most relevant intervention functions to tackle the key domains. The introduction of strategies using the intervention function of Coercion (not currently in use) might not be considered acceptable/appropriate in the hospital context and future research could explore the practicalities of introducing this intervention function in secondary care settings (eg, having behavioural/letter commitments for staff involved in CURE).42 This strategy was successful in avoiding inappropriate antibiotic prescribing by having poster-sized commitment letters featuring clinician photographs and signatures stating a commitment in wards.43
The inclusion of fiscal measures (ie, using the tax system to reduce or increase the financial cost), and legislation (ie, making or changing laws) was considered less practicable in the hospital context. For the policy category of regulation, further strategies could be introduced, for example, establishing rules or principles for vaping within the hospital premises, and further evaluated through research.
The findings presented in this paper are related to the CURE pilot implementation strategy within an acute care setting. Given the long-term plan aims to roll out similar tobacco dependence services within acute, maternity and mental health settings,1 it will be important to conduct qualitative work and strategic behavioural analysis in other contexts where the delivery and/or barriers/facilitators might be different. In addition, suggested future research should also try to understand how these findings differ in different geographical locations given different structures and systems within hospitals. Implementation fidelity across different pilot sites should be evaluated and compared with adherence to protocols. For example, implementation fidelity could be assessed by measuring the completeness of smoking cessation consultation forms and the proportion of patients for whom cessation medications were ordered in hospital.
Conclusion
Despite treating tobacco dependence being one of the most cost-effective health interventions any healthcare system can provide, adherence to smoking cessation standards within hospitals settings remains poor in England. This strategic behavioural analysis study demonstrates how the use of a variety of behaviour change tools can be used to specify the content and possible mechanisms of action of an existing implementation strategy which has achieved some level of success in clinical practice but requires further improvement and evaluation. The CURE implementation strategy may be further optimised by using additional theoretically congruent BCTs to target the less commonly addressed influences related to the social and environmental context (eg, ‘restructuring the physical environment’ by creating a steering group to consider options for discharge pathways).
This study provides comprehensive evidence about current practice in the pilot site that can further inform implementation strategy improvement and the implementation of an NHS-funded tobacco dependence treatment and policy in secondary care in England.
Data availability statement
Data are available upon reasonable request. The data sets used and/or analysed during the current study are available from the corresponding author on reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
Ethical Approval was granted from Northumbria University Faculty of Health and Life Sciences Ethics Committee (Ref 21358). Informed consent was obtained from all study participants. All methods were carried out in accordance with relevant guidelines and regulations. Participants gave informed consent to participate in the study before taking part.
Acknowledgments
We thank all of the healthcare professionals who took the time to participate in this research, the CURE team and PHE Behavioural Insights and Tobacco Control teams for their helpful comments and insight, particularly Martyn Willmore, Aleksandra Herbec, Anna Sallis and Michelle Havill.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Contributors AMR, AH and CH developed the initial study design and secured funding for the study. AW conducted preparation of study materials, data collection and analysis for the qualitative interviews and drafted summary reports. AMR conducted the behavioural analysis. CH developed the logic models. AMR and AW drafted the manuscript. CH, AH, VM, FH and ME contributed and provided comments on data analysis and interpretation, and report drafts. All coauthors have reviewed and agreed the final draft of the paper submitted for publication. AMR is acting as guarantor.
Funding This research was commissioned and funded by Public Health England (award/grant number: not applicable).
Competing interests None declared.
Patient and public involvement Patients and/or the public were not involved in the design, or conduct, or reporting, or dissemination plans of this research.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.