Applying the behavior change wheel to design de-implementation strategies to reduce low-value statin prescription in primary prevention of cardiovascular disease in primary care

Introduction

De-implementation or abandonment of ineffective or low-value healthcare has gained great interest in recent decades, due, among other factors, to the growing empirical evidence of its high prevalence and its impact on patient safety, resource consumption and social inefficiency (1, 2). Accordingly, theoretical and empirical evidence-based knowledge about the key factors (barriers, facilitators, etc.) required for the design and application of de-implementation of low value care, is growing rapidly (2–4).

In light of current understanding of how to tackle targeted implementation and de-implementation endeavors, several fundamental aspects can be concluded. Firstly, factors driving the implementation of both evidence-based and inappropriate interventions in the clinical setting are multi-level, complex, and context specific (2). Therefore, a wide range of factors related to the intervention or practice to be de-implemented, the agents involved in this practice (e.g., healthcare professionals, patients) and other inner and outer context factors, should be carefully accounted for. Secondly, in order to be able to change a certain clinical practice, involved agents and stakeholders should be engaged in the process of identifying the practice determinants, in the design of focused interventions and implementation strategies and in the planning of their assessment (5, 6). Finally, the intervention design should be performed following a formal analysis of the target behavior and its mechanisms of action, and guided by models or theories covering the whole range of potential behavior influences or determinants at stake (7, 8). To this end, Behavioral Science and behavior change theories and frameworks for the development or planning of interventions can aid in better identifying and understanding the multi-level mechanisms that altogether influence clinical behavior, as well as in the selection of focused, effective techniques to promote behavior change of healthcare professionals (9, 10). Some examples of such theories or frameworks for the development or planning of interventions are the PRECEDE-PROCEED model (11), Implementation Mapping (12), or the Behavior Change Wheel (BCW) (13). Despite the wealth of recent scientific literature on the development of intervention or implementation strategies to facilitate the uptake of innovative or evidence-based practices, the application of behavioral science theory for the development of de-implementation interventions is scarce (14–16).

The DE-imFAR (from the Spanish for DE-implementation of low-value pharmacological prescriptions) study aims to carry out a structured, evidence-based and theory-informed process involving the main stakeholders (healthcare managers, professionals, patients, and researchers) for the design, deployment and evaluation of targeted de-implementation strategies for reducing low-value pharmacological prescribing (17). Specifically, the selected low-value practice is the prescription of statins in primary prevention of cardiovascular disease (CVD) in patients with low cardiovascular risk (CVR) (i.e., <10% CVR according to the REGICOR equation). Based on the evidence (18) and clinical practice guidelines (19–21), it is recommended not to start statin therapy in this population, with promotion of healthy lifestyles (i.e., healthy diet, physical activity and smoking cessation) being the recommended intervention instead.

Despite these recommendations, over the last years the consumption of statins in the Basque Country has increased substantially due, in part, to a CVD prevention approach excessively focused on the control of lipid levels and the use of medications. In this regard, the results from a descriptive observational study (as part of the DE-imFAR study) with data from electronic health records (EHR) on the inappropriate prescription rate of statins in patients aged 40–75 years with no history of CVD, with moderately cholesterol but with a CVR <5% (REGICOR) showed an incidence of new inappropriate prescriptions of 10.5 per 100,000 people/year (22). Furthermore, over 60% of the EHRs of these patients with inappropriate prescription of statins did not have a record of having been given advice on physical activity or a healthy diet. Likewise, 49% did not receive preventive recommendation on smoking cessation (22).

This paper reports on Phase I of the DE-imFAR study. Its main aim is to conduct a formative research in the specific context of the Basque Health Service-Osakidetza to: i) understand the problem of low-value statin prescription in primary prevention of CVD and define it in behavioral terms; ii) identify the main determinants of this clinical practice (e.g., at personal, inter-personal, organizational, social level) that must be addressed to change this behavior, and iii) map potential de-implementation strategies; and iv) prioritize mapped strategies based on their perceived effectiveness, feasibility and acceptability.

Materials and methods

Design

A Phase I formative study applying systematic, comprehensive and evidence-based frameworks, such as the Theoretical Domains Framework (TDF) (23, 24) and the BCW (13, 25) for the collaborative design and development of de-implementation strategies to favor the abandonment of low-value pharmacological prescribing of statins in primary prevention of CVD. The DE-imFAR study protocol was reviewed and approved by the Basque Country Clinical Research Ethics Committee (Reference: PI2019102, approved on 10 April 2019) and was registered in the U.S. NLM ClinicalTrials.gov database (ClinicalTrials.gov Identifier NCT04022850, 17 July 2019).

A working group, which was composed of experts in the design of implementation strategies, methodologists, pharmacists, qualitative researchers, clinicians and health service managers, used the TDF and BCW to identify, select, adapt and define possible behavioral change interventions operationalized as de-implementation strategies to address the prioritized determinants of inappropriate statin prescription in CVD primary prevention. This process involved eight steps grouped into the following three stages:

1st stage-understand the behavior: Step 1) define the problem in behavioral terms; step 2) select the target behaviors; step 3) specify the target behaviors; step 4) identify what needs to change

We conducted a set of five semi-structured interviews with a sample of Family Physicians (FPs) (n = 3) and Practice Nurses (n = 2) with recognized expertise and experience in CVD prevention in order to identify the overall behavioral scenario and break down the chain of behaviors and concomitant non-behavioral (e.g., contextual) factors (step 1). The interview script was centered on determining how physicians address and manage the clinical encounters related to CVD prevention, and what the main steps taken are. Three members of the working group independently reviewed the recordings of the interviews, and identified and proposed a set of possible target behaviors. Subsequently, using matrices and exercises proposed by the BCW (25), the working group proceeded to vote and discuss until agreement in order to select (step 2) and to specify (step 3) the final target behaviors most likely to lead to the desired behavior change.

In order to explore the practice determinants (barriers and facilitator) of the selected target behaviors related to inappropriate statin prescription and healthy lifestyle promotion actions, a qualitative study comprising focus groups with FPs was performed (step 4).

Since the DE-imFAR study was carried out in two of the 13 Integrated Healthcare Organizations (IHO) of Osakidetza, a convenience sampling strategy for the recruitment of the healthcare professionals was used. In short, emails were sent to all the FPs from the Ezkerraldea-Enkarterri-Cruces (n = 83) and Barakaldo-Sestao (n = 123) IHOs with a brief explanation of the project and the invitation to participate. Out of the total number contacted, it was possible to recruit 21 FPs. Four focus groups of about 90 min of duration were conducted, two for each IHO, with between four to seven attendees in each group.

The groups were led by two researchers with experience in qualitative research methods, as well as knowledge of the clinical field. The focus groups were audio recorded and transcribed verbatim. Informed consent from all participants was obtained prior to any research procedure. The script of the focus groups were developed to explore in-depth potential determinants with questions covering each of the TDF dimensions (24). An inductive analysis based on grounded theory (26) was adopted to favor the emergence of a theory on the functioning of inappropriate statin prescription based on the words and phrases of the professionals. To facilitate the analysis, a coding scheme regarding the TDF dimensions and their relative constructs was developed. Two researchers independently reviewed and coded the transcripts and iteratively discussed possible discrepancies until reaching a consensus.

In addition, and with a twofold goal of firstly, ascertaining patients' experience regarding the clinical practice of statin prescription; and secondly, of triangulating physicians discourse, a focus group with patients was also conducted. To do so, participating professionals were requested to provide a list of patients “at low cardiovascular risk, in treatment with statins,” as well as for authorization to contact them. Eleven identified patients were contacted by telephone explaining the general objective of the study and the specific purpose of the discussion group and were invited to participate. Finally, a discussion group made up of six patients who agreed to participate was held. Informed consent from all patients was obtained prior the focus groups commencement. The following aspects were explored: how the pharmacological treatment was started; whether it was a decision made in conjunction with the FP; how they were informed; what factors could determine this action (preference or health problem, and at patient, professional, health center level), patient comfort with treatment, and so on.

2nd stage-identify intervention options: Step 5) select intervention functions; step 6) select the specific behavior change techniques

The goal at this stage was to identify the Behavior Change Techniques (BCTs) for each of the agreed determinants of selected target behaviors. Two researchers proceeded first to group each identified barriers and facilitators into their respective TDF domain. Subsequently, they described identified determinants in the form of “what needs to change” and linked them to the intervention functions guided by the BCW instructions and suggestions. Then, all potential policy categories were identified. Lastly, potential BCTs most likely to produce a change were selected for each determinant using the process established by the BCW (25).

3rd stage-identify implementation procedures: Step 7) select strategies and intervention techniques; step 8) select the mode of execution of the intervention

Final definition, packaging and selection of previously identified de-implementation strategies were carried out through a participatory consensus process in the form of round meetings involving the working group as representatives of the main stakeholders. In short, the working group first grouped and logically ordered all related mapped BCTs (i.e., those hypothesized to address the same determinant or several determinants at a time). Then and guided by examples of de-implementation interventions within the literature and by the experience of several team members in the design of implementation strategies, the working group decided upon a clear layout of the techniques to be applied (i.e., the actual content of the interventions, their possible formats and modes of execution) for each of the potential interventions identified through this structured mapping process.

Priorization of identified de-implementation strategies

Finally, in order to prioritize the de-implementation strategies derived from the conducted mapping, a poll process using the LimeSurvey platform involving FPs who collaborated in the focus groups was carried out. Specifically, they assessed the potential effectiveness, acceptability and feasibility of each identified strategy. The prioritization analysis, taking into account the ordinal nature of the measurement scale, was carried out by counting the number (proportion) of observations in each value of the assessed variables. Those considered potentially effective while highly acceptable and feasible for enacting behavior change were prioritized as the final set of specific strategies, to be contained in at least one broad de-implementation strategy seeking to reduce low-value pharmacological prescribing in the primary prevention of CVD.

Results

1st stage-understand the behavior

Steps 1, 2, and 3. Define the problem in behavioral terms, select the target behaviors, and specify the target behaviors

Firstly, derived from the five semi-structures interviews with FPs and Nurses conducted in step 1, the working group defined the target behavior as: “Reduce the prescription of statins in the context of primary prevention of CVD in low-risk patients (REGICOR <5%) and favor the adoption and implementation of the recommended intervention, the promotion of healthy lifestyles (regular physical activity, healthy diet and giving up smoking) at any opportunistic or programmed health center visit for screening or addressing CVD risk factors” (Supplementary Table S1).

Afterwards, in steps 2 and 3, this target behavior was broken down into the chain of behaviors involved and the concomitant precipitating factors (Supplementary Table S2). Three precipitating factors for the practice of primary prevention of CVD were identified: i) alarm systems integrated within the EHR prompting the fulfillment of the Preventive Activities Program (PAPPs); ii) the presence of high cholesterol levels in a blood test result; or, iii) the presence of a prescription initiated or suggested by another professional (specialist or private). Regarding the preventive action behaviors by FPs and Practice Nurses, seven main steps were identified, ranging from the initial general approach for CVD primary prevention focused on CVD risk and the cholesterol level to the enactment of the decided treatment or intervention, the options being the prescription of a statin, the delivery of a healthy lifestyle promotion intervention, or both. The following specific behavior was prioritized by the working group and described according to who needs to do what, when, where, how often, how and with whom, as that most likely to bring about change: “The FP considers options and makes the clinical decision on intervention/treatment to be provided, based on the result of the CVD risk estimation, on knowledge and heuristics in relation to the recommended practice, their attitudes, expectations and abilities, and other contextual factors (time, work overload, organizational norms, decisional fatigue, etc.).”

Step 4. Identify what needs to change

Numerous determinants, facilitators of the inappropriate statin prescription and barriers toward healthy lifestyle promotion emerged from the focus groups with healthcare professionals. Determinants were identified from the quotes guided by a pre-specified coding. Though professionals' discourse tended to saturation, we do not have explicit confirmation of having reached saturation of data with the four groups. Table 1 displays some examples of quotes classified by the domains of the TDF. Except from one TDF dimension, Optimism, all the rest of the dimensions were covered in the FPs' discourses (see Table 1 for extracted quotes):

TABLE 1

Table 1. Quotes extracted from FPs discussion groups by theoretical domains framework determinant dimensions.

Knowledge

Participants felt that lack of awareness of the problem, doubts, clinical guidelines being out of date, and lack of consensus on or variability of recommendations, are main facilitators for an inappropriate prescription (see quotes K_Q1-Q3; Table 1). They believed that clearer evidence and getting a broader vision considering further risk factors would help to prescribe properly (K_Q4, Q5).

Skills

Differential required skills of alternative behaviors, statin prescription versus healthy lifestyle promotion, due to their perceived or experienced ease/difficulty seem to be, on the one hand, a facilitator of an inappropriate prescription, and on the other hand, a barrier to the recommended practice to be provided, especially regarding the prescription of physical activity (Sk_Q1,Q2).

Beliefs about capabilities

The main determinant related to capabilities is the low perceived confidence in prescribing healthy lifestyles, a clinical practice considered difficult in itself as compared to prescribing a statin (Cap_Q1, Q2). This problem is augmented by the difficulties faced by professionals to tackle healthy lifestyle promotion actions as a means of preventing CVD in low-risk patients, who are not usually frequent attenders (Cap_Q3).

Beliefs about consequences

The fear of negative consequences of not treating seemed to be a powerful driver of inappropriate prescribing (Con_Q1). This “defensive medicine” was also enhanced by the perceived effectiveness of statins in decreasing cholesterol levels (Con_Q2). Obtaining such a positive clinical result in the short term contrasted with the long term (and somewhat unperceived) benefits of healthy lifestyle promotion actions (Con_Q3). The adverse effects associated with statins seemed to be a potential barrier to statin prescription (Con_Q4).

Motivation, goals, and intent

The abovementioned scarcity of positive expected results from healthy lifestyle promotion actions has derived in a low motivation of professionals (M_Q1). Actual intention in the form of action plans or goals, both for not prescribing statins and for providing healthy lifestyle promotion interventions, is seen as a necessary condition to endorse guideline-concordant CVD primary prevention efforts (M_Q2).

Memory, attention, and decision-making

A repeated theme in physicians' discourse is the influence of clinical inertia in decision-making favored by contextual factors such as lack of time and heavy workload (MAD_Q1, Q2). Pharmacological prescription is perceived to require less cognitive effort in a saturated clinical practice that leads to decisional fatigue. A defensive medicine mindset is always present when deciding upon treatments (MAD_Q3). Physicians also requested the removal of asterisks in patients' blood test results (i.e., an asterisk is placed alongside cholesterol level when value is greater than or equal to 200 mg/dl) as this visual stimulus induces patients' concerns regarding cholesterol levels (MAD_Q4, Q5). Such markers incite cholesterol-control-focused clinical actions.

Environmental context, resources, and constraints

As previously commented, lack of time and the heavy workload experienced in Primary Care are the main obstacles for prevention efforts (E_Q1, Q2). Physicians also perceived that tools within the EHR are useful but somewhat limited for estimating cardiovascular risk, for reminding and fomenting guideline-concordant CVD primary prevention practice, and for restricting inappropriate statin prescribing (E_Q3, Q4). Teaming up with an involved Practice Nurse in order to share prevention efforts facilitated adequate healthy lifestyle promotion actions in primary prevention of CVD (E_Q5). Lack of external resources inside and outside the clinical setting (i.e., allied healthcare professionals, community resources, etc.) limits the reach of prevention efforts, especially in low-risk young adults, as a non-frequent-user population (E_Q6-Q8).

Social and professional role and identity

Lack of coherence in prescription criteria among the different healthcare professionals (i.e., cardiologists, neurologists and interns in addition to FPs) who attend the same patients dilutes responsible clinical practice (Rol_Q1, Q2). Uncertainties regarding limits in responsibility with respect to healthy lifestyle prescribing and fear of questioning each other's clinical decisions help to maintain inappropriate treatments (Rol_Q3).

Social influences

Patients' lack of awareness together with a perception of low susceptibility and vulnerability regarding cardiovascular risk hamper physicians' primary prevention efforts (SI_Q1, Q2). In contrast, due to the importance given by the media and probably fueled by the pharmaceutical industry, cholesterol is “the bad guy” everybody is worried about and needs to be addressed (SI_Q3). Another ambivalence occurs with healthy lifestyles. On the one hand, the population seems to be more conscious about the overall benefits of healthy behavior. But on the other hand, patients seem to have become so used to the message about the need to change to healthy lifestyles that some prefer to take a “magic” drug in the belief that there is no need to change habits (SI_Q4, Q5). In fact, neither the internal context in the health system which does not prioritize healthy lifestyle promotion practice, nor the external context at societal level influenced by media messages and the economic interests of the pharmaceutical industry targeting cholesterol reduction exclusively, are conducive to good CVD primary prevention practice (SI_Q6-Q8). Professionals also perceive that in certain sectors of the population, such as those with lower socio-economic status, the promotion of healthy lifestyle, although being the recommended practice, is very difficult to implement (SI_Q9, Q10).

Emotion

Mixed emotions were reported by physicians who mainly favor inappropriate prescribing. Professionals must make decisions in an emotional climate marked by uncertainty due to the variability of recommendations and limitations of the Clinical Practice Guidelines (CPG) and fear of consequences of not treating (Em_Q1-Q3). The feeling of pleasing the patient coupled with peace of mind after prescribing statins and obtaining “positive” cholesterol results are factors that seem to weigh substantially on decision making (Em_Q4, Q5). In contrast, positive emotions associated with successful healthy lifestyle changes seen in patients are the only emotional asset that favors continuing the work of promoting healthy lifestyles (Em_Q6).

Behavioral regulation

Professionals complained of a poor quality assessment culture in the healthcare system and of lack of standards and indicators established by the organization to anchor and guide clinical performance (BR_Q1). Data are needed to be able to reflect on performance and to be able to set goals, monitor progress and provide useful feedback, and the lack of access to such data prevents reflection and the establishment of objectives, both of which are seen as necessary to correct the problem of inadequacy in drug prescription (BR_Q2-Q4).

Reinforcement

In addition to the mentioned above in relation to objectives and performance indicators, the results of the evaluations of indicators carried out by the organization do not translate into incentives for professionals, which generates demotivation among those professionals willing to do things right (Re_Q1,Q2).

In addition, we conducted one focus group with six patients in order to triangulate professionals' discourse. We must highlight that the majority of the participants indicated a lack of explanation of the prescribed treatment and their desire to be more involved in the treatment decision. Moreover, they believed that family health history has a lot of weight in the decision and they are concerned about it. They reported that only some professionals recommended healthy lifestyles with or without prescription of statins. When we asked about their preferences for doing physical activity or taking a cholesterol-lowering drug in a context of low CVR, different points of view arose: some preferred physical activity while others preferred to combine exercise and pharmacological treatment. Overall, they were satisfied with taking statins although they preferred not to think about the adverse effects.

2nd stage. Identify intervention options

Steps 5 and 6. Select intervention functions and specific behavior change techniques

Table 2 summarizes the conducted mapping process, linking practice determinants for inappropriate statin prescription (mainly facilitators) and for providing healthy lifestyle promotion interventions (mainly barriers) to intervention functions and policy categories, ending with potential BCTs for attaining the desired target behavior. For example, the lack of awareness among patients regarding the problem of inappropriate pharmacological prescription (Facilitator of the low-value practice) can be addressed through persuasion (Intervention function) and communication actions (Policy category) enacted by techniques focused on providing information about health consequences (BCT) of this low-value practice.

TABLE 2

Table 2. Mapping matrix of potential intervention functions, policy categories and Behavior Change Techniques (BCI) to previously identified determinants of inappropriate statin prescription and healthy lifestyle promotion categorized by Theoretical Domains Framework (TDF) dimensions identified from the qualitative study.

3rd stage. Identify implementation procedures

Steps 7 and 8. Select strategies, intervention techniques, and modes of execution

Through various round meetings, the working group agreed and drew up a list of 14 potential de-implementation strategies with their respective format and techniques of delivery (Table 3). The specified strategies ranged from the optimization of informatics tools in the EHR used in the routine clinical context of CVD prevention, to update or develop clinical guidelines and educational materials on primary prevention of CVD based in evidence, periodic sending of audit and feedback regarding clinical practice indicators or patient mediated interventions. As an example of BCTs grouping into a potential strategy or intervention component, the editing or updating of a CPG put together at least three identified and selected BCTs that may impact Knowledge: A credible source, gives Instruction on how to perform a behavior, and can guide goal setting related to the behavior (see Table 2).

TABLE 3

Table 3. Prioritization of the 14 de-implementation strategies derived from the mapping process.

Prioritization of de-implementation strategies

Lastly, the potential strategies were sent back to all the healthcare professionals involved in the discussion groups and two health managers for their evaluation regarding three dimensions: acceptability, feasibility and potential effectiveness. Thirteen complete evaluations (13/23) were received that allowed the prioritization of the de-implementation strategies (see Table 3).

Discussion

This study aimed to report on the application of a systematic, comprehensive, theory-and evidence-informed framework to design potentially effective and feasible de-implementation strategies to favor the abandonment of low-value pharmacological prescribing in CVD primary prevention in low CVD risk patients (17). Specifically, guided by the TDF and the BCW frameworks (13, 23–25), we have conducted a series of actions to identify determinants of low-value practices and behavioral objectives as areas for improvement, which have helped us to design, operationalize and prioritize various de-implementation strategies.

Avoiding or substituting proven potentially harmful, ineffective or inefficient medical practices is important for improving the quality of healthcare while ensuring sustainability of healthcare systems, which is the reason why in recent years the interest in and the evidence base related to successful de-implementation strategies to favor the abandonment of low-value practices has grown quickly. Statins are among the most widely prescribed medications globally and are increasingly used to prevent CVD in people without CVD (“primary prevention”). However, statins have no or low value for the primary prevention of CVD in low-risk patients (18). On the other hand, healthy lifestyle promotion interventions in clinical settings have been shown to be effective and are the preferred recommended practice, especially in low-risk patients (19–21).

From the growing scientific evidence in implementation research it is known that factors determining the implementation of both evidence-based and inappropriate interventions in the clinical setting are multi-level, complex, and context specific (2). Consequently, the design of interventions should be performed following a process of formal analysis of the target behavior and its theoretically predicted mechanisms of action, all guided by models or theories that cover the entire range of possible influences or determinants of the behavior in question (7–10). Through the performed qualitative study with both main involved healthcare professionals (FPs and Practice Nurses) and affected users (low-CVR-risk patients with inappropriate prescription of statins), we have identified multi-level determinants of the target low-value practice within the context of two IHOs in the Basque Health Service-Osakidetza. Almost all of the dimensions of the TDF have been called into play, as at least one practice determinant (barrier or facilitator) was included in these dimensions. Some of the most consistently reported determinants professionals' focus groups were the lack of time and external resources, preferences and characteristics of patients, limitation of available clinical tools and CPGs, social pressures, fears about negative consequences of not treating high cholesterol levels with drugs, and lack of skills and training of professionals in healthy lifestyle promotion. Patients' main determinants were the lack of explanation of the situation at the medical appointment, the desire to be more involved in the treatment decision, belief and concern about family health history in the decision, and the lack of healthcare professional's recommendation about healthy lifestyles with or without prescription of statins.

The identified determinants are in line with other determinants identified or reported in previous studies regarding determinants of low-value practice and of low-value pharmacological prescription. For example, uncertainty due to the variability and/or conflict of the guidelines with respect to the recommended practice, the pressures and demands on the part of the patients, the need for rapid and decisive action in response to the reasons for consultation and the desire to please the patients have been identified as interconnected motives that generally justify maintaining low-value practices (27). With regard to inappropriate prescription of drugs in general, a systematic review published by Anderson et al. (28), in which the barriers and facilitators for inappropriate prescription were explored, highlights four aspects that facilitate or hinder professionals' decisions when faced with a possible pharmacological prescription: first, awareness of the problem, i.e., knowing to what extent the clinical practice of each professional conforms to what is recommended in CPGs, as well as knowing the consequences of treating a patient pharmacologically or not. Second, self-efficacy, which encompasses the professional's ability to manage the clinical situation based on their knowledge or their ability to offer a non-pharmacological alternative, among others. The third aspect to highlight is inertia, which is a barrier to change in clinical practice; and finally, feasibility, where all the external factors that affect the clinical decision would be included: patient characteristics and preferences, social/cultural factors, prescriptions made by another professional, group pressure and so on. Studies carried out exclusively on the inappropriate prescription of statins emphasized the influence of the perception that professionals have of each patient's CVR and their opinion about the effectiveness and safety of statins (29), beliefs or attitudes toward behavior and perceived control (30); the additional risk factors that the patient may present and the patient's preferences about receiving drug treatment or not (31).

One peculiar aspect in this point is that, due to the addressed clinical scenario (the reduction of low-value prescribing of statins in CVD primary prevention where the promotion of healthy lifestyles is the alternative, recommended practice), this project has attempted to simultaneously identify determinants of both clinical practices. Although it may seem obvious, in such clinical scenarios, stress must be placed on identifying the factors that facilitate or maintain the low-value practice, and on the other hand, the barriers that impede the recommended practice (32).

With the main goal of designing and developing targeted strategies that address the specific determinants of CVD prevention practice in the Basque Health Service-Osakidetza, the present's study main action has been to carry out a mapping process of de-implementation and implementation strategies in order to reduce low-value practices (inappropriate statin prescribing) and promote the implementation of the recommended practice (healthy lifestyle interventions), based on the determinants of routine practice reported by FPs in the focus groups, following the procedure established by the BCW.

The 14 strategies that have emerged from the mapping processes are all “old known” strategies and interventions. Nevertheless, previous studies targeting the reduction of low-value statin prescription have shown some effectiveness of certain dissemination strategies as informative web pages or the implementation of electronic CPGs when compared to routine practice especially when used as multi-component strategies (33–37). Further, educational or training actions for professionals (webinars and workshops), have also shown some effectiveness, especially when combined with other interventions in multi-component strategies (33, 34). And lastly, audit and feedback interventions or those sending a clinical case scenario to professionals (38), and techniques to aid decision-making through clinical decision support systems have achieved good results in increasing the registering of CVR and in adjusting the prescription (39–41).

However, the innovative contribution of having used the BCW is that, both actions, determinant identification and mapping of strategies, aim to target the specific clinical behavior most likely to enable the desired change prioritized by the research group and professionals involved: physicians' decision-making regarding the therapeutic option. Moreover, following a taxonomy of choice architecture techniques (42), all except one of the 14 identified strategies may be categorized as influencing FPs decision-making through three different modes: decision information (e.g., dissemination of CPGs), decision assistance (i.e., alert and reminder systems; involvement of the patient in a shared decision-making process), and decision structure (e.g., audit and feedback system). Furthermore, the agents involved have prioritized the resulting potential de-implementation strategies after assessing their perceived acceptability, feasibility and potential effectiveness. Therefore, though not innovative interventions or strategies, those identified are those that address the specific determinants identified by the protagonists. Research is now needed however, to test whether these barrier-specific strategies for de-implementation identified in the present study are also effective in our context (17).

The present study has several limitations. First, the formative study has been performed in only two IHO of Basque Health Service-Osakidetza that are not representative of all Primary Care centers within our health service. Second, after having invited all professional within the two participating IHOs, only a reduced and auto-selected sample was obtained and we cannot guarantee that we have reached saturation of data regarding physicians discourse related to inappropriate statin prescription. And finally, this previous issue also extends to patients groups by limiting the planned triangulation of discourses to only one group of patients.

Conclusion

The present study aims to contribute to the body of currently scarce literature available on practical de-implementation initiatives by providing detailed illustrations/explanations of our stepped, systematic approach to the design and development of targeted behavior change actions based on prominent available frameworks and theories, mostly from implementation science. Key research questions in implementation science also involve determining what implementation strategies should be provided, to whom, and when, to achieve optimal success in implementing evidence-based clinical practice. As the same paradigm must apply for de-implementation of low-value practices, we propose now to investigate the comparative effectiveness of some/different types or intensities of the prioritized strategies in Phase II of the DE-imFAR project. The future evaluative phase of our study will have the aim of increasing evidence on whether the specific strategies that address determinants of recommended practice in CVD prevention, some similar to those evaluated in the few studies conducted to date, are also effective in our context. If the strategies explored are successful, health planners and managers will have the evidence needed to support the introduction of such structured strategies, informed by the application of methods and procedures of the emerging science of implementation and de-implementation.

Data availability statement

The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

Ethics statement

The studies involving human participants were reviewed and approved by Basque Country Clinical Research Ethics Committee (Reference: PI2019102, approved on 10/04/2019). The patients/participants provided their written informed consent to participate in this study.

Author contributions

AS, JP, and GG conceived the idea, the study guarantors, primarily responsible for the study design and planning, as well as the funding obtained, and will be responsible for project coordination and supervision, analysis and interpretation of results, and manuscript preparation. SP, UE-A, SG-L, MM, IL, RS, ML, OL, and CH collaborated in the study design and were responsible for study coordination, interpretation of results, and manuscript preparation. AS, UE-A, and MM were responsible of the performed analysis. All authors critically reviewed the manuscript and approved this version submitted for publication.

Funding

This project was supported by the Basque Government Department of Health (funded project 2018111085 and 2021111024) and by the Carlos III Institute of Health and co-funded by the European Union (funded project Grant Nos. PI21/00025, RD16/0007/0002, and RD21/0016/0003). The funding bodies had no role in the design of the study, collection, analysis, and interpretation of data or the writing of the manuscript.

Acknowledgments

We thanks are given to the following Family physicians and Practice Nurses that collaborated in the study: Tomás Méndez (Ortuella Primary care Health Center PHC), Idoia González Uribe-Etxebarria (Portugalete PHC), Adelaida Román (Buenavista PHC), Julian Bajo (Kueto PHC), Alberto Arranz (Repélega PHC), Rosana San Emeterio (Buenavista PHC), Dominga González (Buena Vista PHC), Juan Ignacio Urresti (Desierto-Urban PHC), Joaquín Lineo (Rontegui PHC), Josune Aragón (San Vicente PHC), Maite Puente (Zuazo PHC), Aranzazu Azaceta (Zalla PHC), Juana Mª Clemente (Sodupe PHC), Javier Angulo (Sodupe PHC), Javier Amiama (Sodupe PHC), Begoña Ruiz (Sodupe PHC), Juan José Casa (Sodupe PHC), José Antonio Quindimil (Kueto PHC), Fernando Uribe (Desierto-Urban PHC), Sofía García (Zaballa PHC), Karmele Benito del Valle (Lutxana-Barakaldo PHC), Josep Cortada (Deusto PHC), Jesús Torcal (Bsauri-Ariz PHC), Agustín Salazar (Deusto PHC). Acknowledgments and thanks are given to the patients who participated in the study.

Conflict of interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Publisher's note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fmed.2022.967887/full#supplementary-material

References

1. Brownlee S, Chalkidou K, Doust J, Elshaug AG, Glasziou P, Heath I, et al. Evidence for overuse of medical services around the world. Lancet. (2017) 390:156–68. doi: 10.1016/S0140-6736(16)32585-5

PubMed Abstract | CrossRef Full Text | Google Scholar

2. Norton WE, Chambers DA. Unpacking the complexities of de-implementing inappropriate health interventions. Implement Sci. (2020) 15:2. doi: 10.1186/s13012-019-0960-9

PubMed Abstract | CrossRef Full Text | Google Scholar

3. Colla CH, Mainor AJ, Hargreaves C, et al. Interventions aimed at reducing use of low-value health services: a systematic review. Med Care Res Rev. (2017) 74:507–50. doi: 10.1177/1077558716656970

PubMed Abstract | CrossRef Full Text | Google Scholar

4. Grimshaw JM, Patey AM, Kirkham KR, Hall A, Dowling SK, Rodondi N, et al. De-implementing wisely: developing the evidence base to reduce low-value care. BMJ Quality & Safety. (2020) 29:409–17. doi: 10.1136/bmjqs-2019-010060

PubMed Abstract | CrossRef Full Text | Google Scholar

5. Parchman ML, Henrikson NB, Blasi PR, Buist DS, Penfold R, Austin B, et al. Taking action on overuse: creating the culture for change. Healthc (Amst). (2017) 5:199–203. doi: 10.1016/j.hjdsi.2016.10.005

PubMed Abstract | CrossRef Full Text | Google Scholar

6. Willis CD, Saul J, Bevan H, Scheirer MA, Best A, Greenhalgh T, et al. Sustaining organizational culture change in health systems. J Health Organ Manag. (2016) 30:2–30. doi: 10.1108/JHOM-07-2014-0117

PubMed Abstract | CrossRef Full Text | Google Scholar

7. Michie S, Johnston M, Abraham C, Lawton R, Parker D, Walker A, et al. Making psychological theory useful for implementing evidence based practice: a consensus approach. Qual Saf Health Care. (2005) 14:26–33. doi: 10.1136/qshc.2004.011155

PubMed Abstract | CrossRef Full Text | Google Scholar

8. Eccles M, Grimshaw J, Walker A, Johnston M, Pitts N. Changing the behavior of healthcare professionals: the use of theory in promoting the uptake of research findings. J Clin Epidemiol. (2005) 58:107–12. doi: 10.1016/j.jclinepi.2004.09.002

PubMed Abstract | CrossRef Full Text | Google Scholar

9. Atkins L, Francis J, Islam R, O'Connor D, Patey A, Ivers N, et al. A guide to using the theoretical domains framework of behaviour change to investigate implementation problems. Implement Sci. (2017) 12:77. doi: 10.1186/s13012-017-0605-9

PubMed Abstract | CrossRef Full Text

10. Michie S, Richardson M, Johnston M, Abraham C, Francis J, Hardeman W, et al. The behavior change technique taxonomy (v1) of 93 hierarchically clustered techniques: building an international consensus for the reporting of behavior change interventions. Ann Behav Med. (2013) 46:81 −95. doi: 10.1007/s12160-013-9486-6

PubMed Abstract | CrossRef Full Text | Google Scholar

11. Glanz K, Rimer B. Theory at a Glance: A Guide for Health Promotion Practice. 2nd ed. Bethesda: National Cancer Institute; U.S. Department of Health and Human Services; National Institutes of Health (2005).

Google Scholar

12. Fernandez ME, Ten Hoor GA, van Lieshout S, Rodriguez SA, Beidas RS, Parcel G, et al. Implementation mapping: using intervention mapping to develop implementation strategies. Front Public Health. (2019) 7:158. doi: 10.3389/fpubh.2019.00158

PubMed Abstract | CrossRef Full Text | Google Scholar

13. Michie S, van Stralen MM, West R. The behaviour change wheel: a new method for characterising and designing behaviour change interventions. Implement Sci. (2011) 6:42. doi: 10.1186/1748-5908-6-42

PubMed Abstract | CrossRef Full Text

14. Skolarus TA, Hawley ST, Wittmann DA, Forman J, Metreger T, Sparks JB, et al. De-implementation of low value castration for men with prostate cancer: protocol for a theory-based, mixed methods approach to minimizing low value androgen deprivation therapy (DeADT). Implement Sci. (2018) 13:144. doi: 10.1186/s13012-018-0833-7

PubMed Abstract | CrossRef Full Text | Google Scholar

15. Thompson LM, Diaz-Artiga A, Weinstein JR, Handley MA. Designing a behavioral intervention using the COM-B model and the theoretical domains framework to promote gas stove use in rural Guatemala: a formative research study. BMC Public Health. (2018) 18:253. doi: 10.1186/s12889-018-5138-x

PubMed Abstract | CrossRef Full Text | Google Scholar

16. Voorn VM, Marang-van de Mheen PJ, So-Osman C, Vlieland TP, Koopman-van Gemert AW, Nelissen RG, et al. Designing a strategy to implement cost-effective blood transfusion management in elective hip and knee arthroplasties: a study protocol. Implement Sci. (2012) 7:58. doi: 10.1186/1748-5908-7-58

PubMed Abstract | CrossRef Full Text | Google Scholar

17. Sanchez A, Pijoan JI, Pablo S, Mediavilla M, de Rozas RS, Lekue I, et al. Addressing low-value pharmacological prescribing in primary prevention of CVD through a structured evidence-based and theory-informed process for the design and testing of de-implementation strategies: the DE-imFAR study. Implement Sci. (2020) 15:8. doi: 10.1186/s13012-020-0966-3

PubMed Abstract | CrossRef Full Text | Google Scholar

18. Byrne P, Cullinan J, Smith A, Smith SM. Statins for the primary prevention of cardiovascular disease: an overview of systematic reviews. BMJ Open. (2019) 9:e023085. doi: 10.1136/bmjopen-2018-023085

PubMed Abstract | CrossRef Full Text | Google Scholar

19. San Vicente Blanco R., Pérez Irazusta I., Ibarra Amarica J., Berraondo Zabalegu2i I., Uribe Oyarbide.F., Urraca Garcia de Madinabeitia J., et al. Guía de Práctica Clínica sobre el manejo de los lípidos como factor de riesgo cardiovascular. Osakidetza. Vitoria-Gasteiz. Available online at: https://portal.guiasalud.es/wp-content/uploads/2018/12/GPC_567_Lipidos_Osteba_compl.pdf.

Google Scholar

20. National Institute for Health and Care Excellence. Cardiovascular Disease: Risk Assessment and Reduction, Including Lipid Modification [Internet]. London: NICE (Clinical guideline [CG181]) (2014). Available online at: https://www.nice.org.uk/guidance/cg181

21. Grundy SM, Stone NJ, Bailey AL, Beam C, Birtcher KK, Blumenthal RS, et al. 2018 AHA/ACC/AACVPR/AAPA/ABC/ACPM/ADA/AGS/APhA/ASPC/NLA/PCNA guideline on the management of blood cholesterol: a report of the American College of Cardiology/American Heart Association Task Force on Clinical Practice Guidelines. Circulation. (2019) 139:e1082–e1143. doi: 10.1161/CIR.0000000000000624

PubMed Abstract | CrossRef Full Text | Google Scholar

22. Elizondo-Alzola U, Sánchez A, Pijoan JI, Mediavilla MM, Sainz De Rozas R, Lekue I, et al. Statins in primary prevention of cardiovascular disease: incidence of potentially inappropriate prescriptions in very low risk primary care patients and associated factors. J Gen Prac. (2022) 10:456. doi: 10.37421/2329-9126.22.10.461

CrossRef Full Text

23. French SD, Green SE, O'Connor DA, McKenzie JE, Francis JJ, Michie S, et al. Developing theory-informed behaviour change interventions to implement evidence into practice: a systematic approach using the Theoretical Domains Framework. Implement Sci. (2012) 7:38. doi: 10.1186/1748-5908-7-38

PubMed Abstract | CrossRef Full Text

24. Cane J, O'Connor D, Michie S. Validation of the theoretical domains framework for use in behaviour change and implementation research. Implement Sci. (2012) 7:37. doi: 10.1186/1748-5908-7-37

PubMed Abstract | CrossRef Full Text

25. Michie S, Atkins L, West R. The Behaviour Change Wheel: A Guide to Designing Interventions. London: Silverback Publishing (2014).

PubMed Abstract | Google Scholar

26. Strauss A, Corbin J. Grounded theory methodology: an overview. In:Denzin N, Lincoln Y, editors. 1st ed. Handbook of Qualitative Research. Thousand Oaks, CA: SAGE (1994). p. 273.

Google Scholar

27. Ingvarsson S, Augustsson H, Hasson H, Nilsen P, von Thiele Schwarz U, von Knorring M. Why do they do it? A grounded theory study of the use of low-value care among primary health care physicians. Implement Sci. (2020) 15:93. doi: 10.1186/s13012-020-01052-5

PubMed Abstract | CrossRef Full Text | Google Scholar

28. Anderson K, Stowasser D, Freeman C, Scott I. Prescriber barriers and enablers to minimizing potentially inappropriate medications in adults: a systematic review and thematic synthesis. BMJ Open. (2014) 4:e006544. doi: 10.1136/bmjopen-2014-006544

PubMed Abstract | CrossRef Full Text | Google Scholar

29. Bonner C, Jansen J, McKinn S, Irwig L, Doust J, Glasziou P, et al. How do general practitioners and patients make decisions about cardiovascular disease risk? Health Psychol. (2015) 34:253–61. doi: 10.1037/hea0000122

PubMed Abstract | CrossRef Full Text | Google Scholar

30. Rashidian A, Russell I. Intentions and statins prescribing: can the theory of planned behaviour explain physician behaviour in following guideline recommendations? J Eval Clin Pract. (2011) 17:749–57. doi: 10.1111/j.1365-2753.2011.01690.x

PubMed Abstract | CrossRef Full Text | Google Scholar

31. Clough JD, Martin SS, Navar AM, Lin L, Hardy NC, Rogers U, et al. Association of primary care providers' beliefs of statins for primary prevention and statin prescription. J Am Heart Assoc. (2019) 8:e010241. doi: 10.1161/JAHA.118.010241

PubMed Abstract | CrossRef Full Text | Google Scholar

32. Davidson KW, Ye S, Mensah GA. Commentary: de-implementation science: a virtuous cycle of ceasing and desisting low-value care before implementing new high value care. Ethn Dis. (2017) 27:463–8. doi: 10.18865/ed.27.4.463

PubMed Abstract | CrossRef Full Text | Google Scholar

33. Etxeberria A, Alcorta I, Pérez I, Emparanza JI, Ruiz de Velasco E, Iglesias MT, et al. Results from the CLUES study: a cluster randomized trial for the evaluation of cardiovascular guideline implementation in primary care in Spain. BMC Health Serv Res. (2018) 18:93. doi: 10.1186/s12913-018-2863-x

PubMed Abstract | CrossRef Full Text | Google Scholar

34. Keller H, Hirsch O, Kaufmann-Kolle P, Krones T, Becker A, Sönnichsen AC, et al. Evaluating an implementation strategy in cardiovascular prevention to improve prescribing of statins in Germany: an intention to treat analysis. BMC Public Health. (2013) 13:623. doi: 10.1186/1471-2458-13-623

PubMed Abstract | CrossRef Full Text | Google Scholar

35. Keller H, Krones T, Becker A, Hirsch O, Sönnichsen AC, Popert U, et al. Arriba: effects of an educational intervention on prescribing behaviour in prevention of CVD in general practice. Eur J Prev Cardiol. (2012) 19:322–9. doi: 10.1177/1741826711404502

PubMed Abstract | CrossRef Full Text | Google Scholar

36. Comin E, Catalán-Ramos A, M Verdu J, Iglesias-Rodal M, del Val García JL, Grau M, et al. Impact of the implementation of electronic guidelines for cardiovascular prevention in primary care: study protocol. Inform Prim Care. (2012) 20:129–39. doi: 10.14236/jhi.v20i2.33

PubMed Abstract | CrossRef Full Text | Google Scholar

37. Comin E, Catalan-Ramos A, Iglesias-Rodal M, Grau M, Del Val JL, Consola A, et al. Impact of implementing electronic clinical practice guidelines for the diagnosis, control and treatment of cardiovascular risk factors: a pre-post controlled study. Aten Primaria. (2017) 49:389–98. doi: 10.1016/j.aprim.2016.11.007

PubMed Abstract | CrossRef Full Text | Google Scholar

38. Dormuth CR, Carney G, Taylor S, Bassett K, Maclure M. A randomized trial assessing the impact of a personal printed feedback portrait on statin prescribing in primary care. J Contin Educ Health Prof. (2012) 32:153–62. doi: 10.1002/chp.21140

PubMed Abstract | CrossRef Full Text | Google Scholar

39. Peiris D, Usherwood T, Panaretto K, Harris M, Hunt J, Patel B, et al. The Treatment of cardiovascular Risk in Primary care using Electronic Decision supOrt (TORPEDO) study-intervention development and protocol for a cluster randomised, controlled trial of an electronic decision support and quality improvement intervention in Australian primary healthcare. BMJ Open. (2012) 2:e002177. doi: 10.1136/bmjopen-2012-002177

PubMed Abstract | CrossRef Full Text | Google Scholar

40. Peiris D, Usherwood T, Panaretto K, Harris M, Hunt J, Redfern J, et al. Effect of a computer-guided, quality improvement program for cardiovascular disease risk management in primary health care: the treatment of cardiovascular risk using electronic decision support cluster-randomized trial. Circ Cardiovasc Qual Outcomes. (2015) 8:87–95. doi: 10.1161/CIRCOUTCOMES.114.001235

PubMed Abstract | CrossRef Full Text | Google Scholar

41. Patel B, Peiris D, Usherwood T, Li Q, Harris M, Panaretto K, et al. Impact of sustained use of a multifaceted computerized quality improvement intervention for cardiovascular disease management in australian primary health care. J Am Heart Assoc. (2017) 6:e007093. doi: 10.1161/JAHA.117.007093

PubMed Abstract | CrossRef Full Text | Google Scholar

42. Münscher R, Vetter M, Scheuerle T. A review and taxonomy of choice architecture techniques. J Behav Decis Mak. (2016) 29:511–24. doi: 10.1002/bdm.1897

CrossRef Full Text | Google Scholar