Mapping implementation barriers and facilitators of a physical activity consultation in the Portuguese primary healthcare: a qualitative approach combining the Tailored Implementation for Chronic Diseases and Behaviour Change Wheel frameworks Authors: Catarina Santos Silva, Marlene N Silva, Jorge Encantado, Sofia Franco, Bruno Avelar Rosa, Romeu Mendes, Pedro J Teixeira, Cristina A Godinho Published: 2025 ABSTRACT AbstractObjectiveDespite evidence of the cost-effectiveness of physical activity (PA) promotion interventions in healthcare settings, translating them into practice remains challenging. This study aimed to identify implementation barriers and facilitators of a Portuguese PA consultation programme implemented in primary healthcare of the Portuguese National Health Service. Additionally, it sought to inform future implementation strategies, using a theoretically based approach.DesignQualitative interview study, using both deductive and inductive approaches.SettingPrimary healthcare units across all health administration regions of mainland Portugal.ParticipantsTwenty-eight participants (six medical doctors, five exercise professionals and 17 patients) from all health regions of the country, involved in the implementation of the Portuguese PA prescription consultation.ResultsFifty-three categories of determinants were identified, using the Tailored Implementation for Chronic Diseases framework. Key barriers included ineffective referral processes to the consultation, challenges in integrating the intervention with existing healthcare demands and insufficient local/regional prioritisation of PA promotion. Key facilitators included high intervention acceptability, diverse community PA resources and good interpersonal skills of implementers. Drawing on the Behaviour Change Wheel, theoretically based inputs to design strategies addressing each barrier were provided.ConclusionsThe implementation of PA consultation was influenced by a broad range of determinants. The most frequently reported barriers are primarily structural and opportunity-related, suggesting system-level implementation strategies are most appropriate. Future strategies should consider implementing clinical standards/orientations for PA promotion, providing institutional incentives based on the attainment of PA indicators, expanding consultation coverage and diversifying referral strategies, reinforcing health system-community partnerships and strengthening training opportunities for implementers. These findings offer relevant insights for enhancing the future implementation of PA consultations, for scaling them up and, ultimately, to increase their effectiveness. STRENGTHS AND LIMITATIONS OF THIS STUDY The use of validated frameworks (Tailored Implementation for Chronic Diseases; Capability, Opportunity, Motivation-Behaviour model; and Behaviour Change Wheel) contributed to providing a holistic understanding of implementation determinants. The use of validated frameworks (Tailored Implementation for Chronic Diseases; Capability, Opportunity, Motivation-Behaviour model; and Behaviour Change Wheel) contributed to providing a holistic understanding of implementation determinants. Data collection from both deliverers and receivers (ie, medical doctors, exercise professionals and patients) contributed to capturing the diversity of existing perspectives. Data collection from both deliverers and receivers (ie, medical doctors, exercise professionals and patients) contributed to capturing the diversity of existing perspectives. Assessing implementation determinants based on frequency of report may not totally reflect the relative relevance of each factor. Assessing implementation determinants based on frequency of report may not totally reflect the relative relevance of each factor. The retrospective and self-reported nature of data collection may introduce recall and social desirability bias, potentially leading to the omission of relevant factors. The retrospective and self-reported nature of data collection may introduce recall and social desirability bias, potentially leading to the omission of relevant factors. Women represented only one-third of the sample and patients from one of the country’s Health Regions were not possible to recruit. Women represented only one-third of the sample and patients from one of the country’s Health Regions were not possible to recruit. Introduction Promoting physical activity (PA) in healthcare settings is cost-effective against physical inactivity,1 a significant risk factor for non-communicable diseases and premature deaths.2 PA counselling and prescription are well-suited to primary care3 and strongly endorsed by international authorities, such as the World Health Organization (WHO) and the Organisation for Economic Cooperation and Development.1 4 5 Primary healthcare offers broad population access across their life course, continuity of care and opportunities to address PA and potential links to community resources.6 7 Medical doctors are viewed as pivotal agents in encouraging patients to increase PA. They are considered a credible and trusted source, with ongoing patient contact, capable of influencing patients’ behaviours.8 Despite growing evidence supporting the effectiveness of primary care PA counselling and prescription under controlled conditions in improving patients’ PA levels,911 translating them into practice poses significant challenges.1214 In 2021, only 40% of countries had implemented a national protocol for PA promotion in healthcare settings,15 despite WHO’s Global Action Plan for Physical Activity 2018–2030 identifying this as a priority.4 In Portugal, the Directorate-General of Health under the Portuguese Ministry of Health introduced a national model in 2017 for promoting PA that included—besides PA assessment and counselling—the creation of a consultation solely dedicated to PA prescription and referral.16 The intervention components are evidence-based and align with the core elements of the Swedish Physical Activity on Prescription model, including (a) person-centred individualised counselling, (b) written prescriptions, (c) evidence-based PA recommendation, (d) follow-up and (e) supporting environment, community-based network.5 Piloted in the National Health Service, the PA consultation faced implementation challenges, underscoring the need for context-specific assessments of barriers and facilitators for better implementation. Understanding contextual barriers and facilitators has been recommended, due to their critical role in developing tailored implementation strategies aimed at refining interventions’ implementation.1214 Implementation science emphasises this process to bridge the gap between research and practice. Barriers to PA prescription include individual factors (eg, health professionals’ knowledge or trust in referring patients to external resources), organisational issues (eg, competing time demands and challenges in coordinating with other sectors) and sociopolitical obstacles (eg, lack of financial incentives). However, few studies have applied framework-based methods to identify these determinants,17 which could provide a more comprehensive and systematic understanding of implementation context and better support applied health policy research specific aims.18 Moreover, there is limited implementation research on consultations dedicated exclusively to PA prescription and referral. Also, since some implementation barriers and facilitators can be context-specific and previous studies have focused on other countries, it is important to further investigate whether certain implementation determinants are applicable across different contexts or remain context-dependent. This study specifically aimed to (a) identify implementation barriers and facilitators of the Portuguese PA prescription consultation and (b) explore future implementation strategies for improving these consultations and supporting the long-term maintenance and scaling-up of this and other similar interventions, using a theoretically based approach. Methods This study adopted a qualitative descriptive approach19 and was reported in accordance with the Standards for Reporting Qualitative Research20 (see online supplemental table 1). Data presented was collected in 2019, during the first implementation phase of the PA consultation. Study approach This study used a qualitative approach for identifying implementation barriers and facilitators of the PA consultation. Semi-structured interviews were conducted with both deliverers and receivers (ie, medical doctors, exercise professionals and patients) on their experiences and perspectives with the PA consultation implementation. Within a postpositivism paradigm,21 both deductive (framework-driven) and inductive (considering new emerging data not reflected in the framework) approaches were employed. Two frameworks were used, according to study goals. To identify implementation barriers and facilitators of the Portuguese PA prescription consultation (study goal I), the Tailored Implementation for Chronic Diseases (TICD) framework was used. The TICD framework is particularly notable for its comprehensive approach and for its focus on healthcare contexts and professional practices. It constitutes a screening tool for identifying determinants of practice and is intended to facilitate implementation studies and quality improvement initiatives. Based on 12 checklists, it comprises 57 determinants across seven domains: guideline factors, individual health professional factors, patient factors, professional interactions, incentives and resources, capacity for organisational change, and social, political and legal factors.22 To explore future implementation strategies (study goal II), the Behaviour Change Wheel (BCW) was applied. The BCW links behaviour change determinants to intervention options and policies, guiding how to shape and improve implementation behaviours. It offers a practical method to select appropriate intervention and policies based on identified behavioural/implementation determinants. It comprises three layers: (1) the ‘COM-B system’ at the core, composed of three interacting conditions necessary for a behaviour to occur: Capability (physical and psychological), Opportunity (physical and social) and Motivation (reflective and automatic); (2) nine intervention functions, around the COM-B axis: education, persuasion, incentivisation, coercion, training, restriction, environmental restructuring, modelling and enablement; (3) seven policy categories, around the intervention functions (a policy corresponds to any action from authorities that influences interventions’ occurrence): communication/marketing, guidelines, fiscal, regulation, legislation, environmental/social planning and service provision. Definitions of intervention and policy types are provided by Michie et al.23 Researcher characteristics and reflexivity The first author is a female researcher, with a background in nutrition sciences and PA. She has 9 years of working experience in trials with mixed-methods evaluation approaches, having had mentoring and observational trainings in qualitative methodologies. Other authors’ backgrounds are psychology, PA and nutrition sciences, all having expertise in qualitative research. The authors were members of the National Physical Activity Promotion Programme team at the Directorate-General of Health when this study was conducted. This work is part of a PhD in human kinetics, focused on PA and health. The Portuguese PA consultation The implementation of the PA consultation involved 14 health units nationwide,16 from the five Administrative Health Regions of mainland Portugal. The implementation was overseen by the Directorate-General of Health and comprised 27 implementers (14 medical doctors, 13 exercise professionals) and 217 patients. Targeting insufficiently active patients with chronic conditions (type 2 diabetes and/or depression, for pilot purposes), eligible patients were referred by their family doctor to a three to four face-to-face consultations programme with 6 months duration (figure 1). It included evidence-based guidelines,24 patient-centred counselling incorporating behaviour change techniques25 and a PA prescription, which can involve a referral and community-based support. It was implemented by a multidisciplinary team—a medical doctor graduated in Sports Medicine and an exercise professional, preferably with a master’s degree in exercise and health. Exercise professionals were sourced from local municipalities or universities, while medical doctors were specialist or intern family doctors within the health units, with an endorsed reduction in usual care to accommodate the PA consultation. All professionals attended training sessions (8 hours for medical doctors, 16 hours for exercise professionals) before implementation, covering PA prescription in type 2 diabetes and depression, behaviour change techniques, consultation protocols and local implementation planning. The Portuguese physical activity (PA) consultation. The meaning of the asterisk (*) is provided in the figure artwork. #When patients receive an individual PA prescription only. Mo, month(s). In each consultation, patients’ eligibility and health status were assessed by the medical doctor, followed by a tailored PA prescription by the exercise professional (individualised exercise or a referral to a community programme). Patients could also be referred to other primary care professionals as necessary (nutritionist, psychologist, physiotherapist, etc). Follow-up calls were made by exercise professionals to assess whether adjustments to the prescription were needed. Implementation was monitored with weekly calls by the research team, to support local implementation. The pilot study included process evaluation of the PA consultation implementation (barriers and facilitators) and outcome evaluation of its impact on patients PA levels and health status. Therefore, patients underwent some assessment procedures on health indicators before each consultation at the health unit that were not related to the aims of the present study. Implemented in 2019, the pilot project was suspended in 2020 due to COVID-19 and resumed 2 years later. Sampling A purposeful sampling of implementers (medical doctors and exercise professionals) was conducted, according to the following procedure: (a) all medical doctors and exercise professionals from participant health units from all the five geographic health administration regions with a minimum 6-month implementation period were invited; (b) invitations were performed by phone call and all aims and procedures of the study were presented and questions clarified. To ensure a comprehensive understanding of the implementation context, perspectives from receivers (patients) were also collected. Patients were identified through a snowball sampling technique,26 according to the following procedure: (a) medical doctors and exercise professionals who agreed to be interviewed were invited to identify patients that they considered to be information-rich cases, capable of providing detailed insights into their experience and who had received the intervention for at least 3 months in their health unit; (b) patients were then approached by the medical doctors/exercise professionals and agreed to be contacted by the research team; (c) the research team contacted the identified patients by phone call to present all study aims and procedures, invite them to participate and clarify any questions regarding the study. Participation was voluntary (no incentives provided). The final sample size was determined by data saturation.26 We used a purposeful sampling with maximum variation approach (eg, diverse geographic regions; diverse roles in the implementation, such as implementers and receivers; ages; habilitations; etc), to access different perspectives on the same subject and gain a comprehensive understanding of the global implementation phenomenon. Data collection Individual semi-structured interviews were conducted in the local language, from 28 July 2020 to 26 July 2021. Due to COVID-19 restrictions, medical doctors and exercise professionals were interviewed via videoconference. Patients with sufficient digital skills also used videoconference, while others were interviewed by phone (n=9). We considered a videoconference format more suitable for individual interviews than for focus groups, given that for participants with lower digital literacy, phone-based focus groups would not be technically feasible. Individual interviews would also allow for more in-depth exploration of personal experiences and perspectives and allow individual participants more privacy in their responses. Each interview was audio-recorded. CSS and JE, both members of the National Physical Activity Promotion Programme but not directly involved in the consultation implementation, conducted the interviews. Topic guides, developed from weekly monitoring notes and the TICD framework,22 explored nine theoretical domains, including recruitment, consultation model, implementation experiences, implementation fidelity (deliverers’ interviews), supporting materials and resources (deliverers’ interviews); results; COVID-19 interruption and future perspectives; intervention sustainability (deliverers’ interviews) and behaviour change maintenance (patients’ interviews) (online supplemental table 2). Sociodemographic information, including age, pathology group (for patients), years of professional experience—both total and specific to implementing PA promotion interventions (for implementers), was also collected for sample characterisation purposes. Implementers verbally provided this information at the end of the interviews, while patients characteristics were retrieved from self-reported questionnaires in the pilot study. No adaptations to the interview guides were made during the interview period. Additionally, written records on recruitment, barriers, facilitators and local solutions were kept during weekly monitoring. Data analysis Recordings were transcribed verbatim using NVivo Transcription software, with accuracy checking against the audio by SF. NVivo Windows (Release 1.7.1) and Excel 365 were used for coding and data analysis. Content data analysis followed a framework approach,18 27 primarily deductive via the TICD framework,22 supplemented by an inductive approach to capture new insights. The analysis comprised five phases18 27: Data familiarisation and initial notes: this phase was conducted by CSS and JE. Both coders read all interview transcripts and took detailed notes on six of them to immerse themselves in the data, focusing on key sections relevant to the study’s aims. Notes were taken only for a subset of the transcripts because coders were also the interviewers and were already familiar with the overall content of the interviews. Data familiarisation and initial notes: this phase was conducted by CSS and JE. Both coders read all interview transcripts and took detailed notes on six of them to immerse themselves in the data, focusing on key sections relevant to the study’s aims. Notes were taken only for a subset of the transcripts because coders were also the interviewers and were already familiar with the overall content of the interviews. Thematic codebook development: using insights from the familiarisation phase and the TICD framework,22 a thematic codebook was developed and first tested on a few transcripts. CSS and JE independently coded one interview to discuss discrepancies and refined the codebook iteratively. A further 10% of data were coded and inter-rater agreement obtained (Cohen’s κ=0.694). As the agreement achieved was substantial and due to the large dataset and limited human resources, the research team deemed a 10% data subset sufficient for agreement calculation. To enhance confidence, regular in-between meetings were also held with the research team (MNS and CG) to refine the codebook during its development. Thematic codebook development: using insights from the familiarisation phase and the TICD framework,22 a thematic codebook was developed and first tested on a few transcripts. CSS and JE independently coded one interview to discuss discrepancies and refined the codebook iteratively. A further 10% of data were coded and inter-rater agreement obtained (Cohen’s κ=0.694). As the agreement achieved was substantial and due to the large dataset and limited human resources, the research team deemed a 10% data subset sufficient for agreement calculation. To enhance confidence, regular in-between meetings were also held with the research team (MNS and CG) to refine the codebook during its development. Coding: CSS systematically applied the refined TICD framework index and codebook (online supplemental table 3) to all transcripts. Since a substantial inter-rater agreement was previously established (Cohen’s κ=0.694)28 and due to limited human resources, this phase was conducted by a single coder, although having, at least, two independent coders would strengthen the process. To enhance confidence in results, two additional researchers (CG and MNS) systematically cross-examined the coded data to ensure coherence, supported by regular discussions with the entire research team. A final verification of all coded data was also conducted at the end of the coding phase by CG and MNS. Coding: CSS systematically applied the refined TICD framework index and codebook (online supplemental table 3) to all transcripts. Since a substantial inter-rater agreement was previously established (Cohen’s κ=0.694)28 and due to limited human resources, this phase was conducted by a single coder, although having, at least, two independent coders would strengthen the process. To enhance confidence in results, two additional researchers (CG and MNS) systematically cross-examined the coded data to ensure coherence, supported by regular discussions with the entire research team. A final verification of all coded data was also conducted at the end of the coding phase by CG and MNS. Charting: CSS rearranged and synthesised data into tables by interviewee type, main themes, subthemes/determinants and frequency of occurrence. The data were then organised by study units (participants) to determine the frequency of participants reporting each barrier and facilitator. Summary statements and illustrative quotes were generated and selected for each code/determinant. Charting: CSS rearranged and synthesised data into tables by interviewee type, main themes, subthemes/determinants and frequency of occurrence. The data were then organised by study units (participants) to determine the frequency of participants reporting each barrier and facilitator. Summary statements and illustrative quotes were generated and selected for each code/determinant. Mapping and interpretation: The barriers and facilitators reported most frequently were mapped against the theoretical constructs of the COM-B model and the BCW23 to frame the results within an implementation theoretical framework. This mapping, conducted by CSS, linked identified determinants to intervention functions and policies, guiding future implementation strategies. To enhance confidence, a final verification of this stage was also conducted by CG and MNS, experts in behaviour change theories. Mapping and interpretation: The barriers and facilitators reported most frequently were mapped against the theoretical constructs of the COM-B model and the BCW23 to frame the results within an implementation theoretical framework. This mapping, conducted by CSS, linked identified determinants to intervention functions and policies, guiding future implementation strategies. To enhance confidence, a final verification of this stage was also conducted by CG and MNS, experts in behaviour change theories. Weekly monitoring records were also reviewed for data triangulation. Sociodemographic data was processed using descriptive statistics. Results Participants Twenty-eight participants were recruited for individual semi-structured interviews from six health units: six medical doctors (mean age=32.8, SD=3.5, two females), five exercise professionals (mean age=39.4, SD=6.9, one female) and 17 patients (mean age=56.5, SD=8.5, 10 females). One invited exercise professional was unavailable due to professional commitments. Table 1 summarises participants’ characteristics. Each health unit had at least one medical doctor, one exercise professional and one patient enrolled, except for two units (one missing an exercise professional and the other missing a patient representative). Due to the pandemic and participants’ preferences, 19 interviews were conducted via videoconference and 9 by telephone. The interviews lasted an average of 40 min and 37 s. Table 1Participants’ characteristicsMedical doctorsExercise professionalsPatientsFemales (%)33.320.058.8Average age (SD)32.8 (3.5)39.4 (6.9)56.5 (8.5)Regional health administration (%) Alentejo16.720.011.8 Algarve16.720.00.0 Centre33.320.052.9 Lisbon and Tagus Valley16.720.023.5 North16.720.011.8Average years of professional experience (SD)7.2 (1.9)17.0 (7.7) Average years of professional experience in PA promotion interventions (SD)2.8 (1.8)11.6 (8.1) Pathology group (%) Depression 64.7 Type 2 diabetes 35.3PA, physical activity; SD, Standard Deviation. Participants’ characteristics PA, physical activity; SD, Standard Deviation. Goal I: key implementation determinants, according to COM-B components Fifty-three categories of determinants were identified, comprising 47 barriers and 52 facilitators, with no new categories from weekly monitoring. Table 2 lists the most frequently reported determinants. Table 2Most frequent implementation determinants of the PA consultationLevel of relevanceImplementation determinantsDescriptive summary sentencesSample quotesFacilitatorBarrierReported by allIntervention factors—intervention components, characteristics, or contentThe multidisciplinarity nature of the consultation was highly valued (and should perhaps involve other specialties)“The specific component of multidisciplinarity with exercise professionals was basically… It is one of the strenghts”. (MD1)Tailored/Patient-centred intervention was highly valued“The customization is excellent. I think this makes a lot of difference in the whole process, because… ehhh… there are different people, different rhythms, different ways of… of… of understanding and of… of wanting. And also time to… availability to do it. So, this… this idea… of personalizing this, of being able to see who is more suited to certain types of activities is… well, it’s interesting”. (P7)The consultations’ length was adequate for purposes“Perfectly adequate, it served perfectly, we were even able to do the initial physical assessment with the STEP test and all that, it was perfectly suited to do this and the counselling and guidance… I think the time was well planned, well stipulated”. (MD4)The consultations’ frequency was considered adequateThe consultations’ frequency lacks having biannual long-term follow-up“0-3-6 months seems fine to me… it might make sense, 0–3, 6 and then every 6… that might make sense, but initially I think 0–3 makes sense, that is, don’t wait 6 months because we will lose the patient halfway through for sure. If we do 0-3-6 and then… imagine this is in the long term—every 6 months, then I think it will make sense”. (MD1)Follow-up phone calls were considered useful“Fundamental… Fundamental, because people cannot be just an email away or even an SMS away, so there has to be much closer contact there. Apart from face-to-face contact, telephone contact… and even when people didn’t answer, we tried again later… and, therefore, this connection is essential…” (EP1)Comprehensiveness of consultation’s assessments was highly appreciated“I thought it was good that the person always had a doctor to take measurements and see if everything was ok with our body… And yes… I thought the model was well made”. (P15)Having both the possibilities (referral to community PA programmes and/or have an individual PA prescription) to offer to patients is pivotal“without a doubt, I think it is an added value, and our physical activity consultation would not make sense without these 2 options, because I cannot impose an individual plan to a person if he/she tells me that they are not going to do it. And I’m not going anywhere, am I?” (MD6)Consultation team based on a mentee format could increase sustainability, as it is not dependent on a restricted teamConsultation implementation and sustainability dependent on a restricted team“(…) professionals who are affected [to the PA consultation], these professionals should have other people to help, like being tutors of these people who also help”. (EP4)Global satisfaction with PA consultation model“I was very pleased that there was such a different initiative, with an approach… I really liked it, I thought it was really good to have this initiative and to have this type of consultation, I think it’s needed”. (P8)Reported by all professionals and very frequent among patientsIntervention factors—referral strategyPatients accept well referrals made by their family doctor“(…) I liked, right away, that the doctor said to me that these PA consultations existed…” (P5)Lack of referral to the PA consultation (by other clinicians)“Literally no one reached me, except the patients referred by me (…)”. (MD2)Facilitating strategies to increase referral to the PA consultation“In a second moment, just with medical doctors, we explained what we wanted, what population was eligible… I left a reminder in front of each colleague’s computer with the inclusion and exclusion criteria for the groups, and, weekly, and whenever the technicians were also present at the Unit, I remember them ‘Look, we need people here for the… remember that this option exists, that this resource is available, all the people who need it and meet the inclusion criteria, remember to refer through SClinico, don’t forget that this exists’”. (MD4)Reported by all professionals and very frequent among patientsIncentives and resources—PA opportunities (constraints)PA resources’ diversity is fundamental to adjusting options to patients’ characteristics“As well as the resources to which we refer patients, because this is also very important. I cannot say to all the patients who come to my consultation ‘you are now going to walk’ and I prescribe walking to everyone, so even my colleagues say ‘For that, you don’t even need to go there for any consultation, I’ll tell you and you’ll be on your way’. Therefore, when we are unable to offer a differentiating response, people say ‘no, don’t go to that consultation because it’s not worth it’… When we are unable to give that response, the consultation loses the value”. (MD3)Sometimes, community PA programmes’ schedules are incompatible with working hours“(…) contrary to the flexibility of the healthcare team, the [community] partner does not have much flexibility and the schedules were not always the most compatible, for the activities I wanted to attend… Because the proposal was to do three activities per week, or three days a week, and given my schedule and the schedules they proposed, I could only do two activities, even though I was signed up for three”. (P9)Reported by all professionalsIndividual health professional factors—skills of interpersonal behaviourIt is important that professionals behave in such a way that patients feel supported, motivated and that professionals really care about their situation“(…) I was quite surprised by the time the team spent with me (…) Not only on a medical level, they paid a lot of attention to everything, to all health details. When I complained about my knee, they never forget afterwards, ‘what about your knee?’ (…) There was always that encouragement, they always went into detail to see blood pressure, which is something I have, high blood pressure … I was very well accompanied”. (P14)It is important that the professionals give clear instructions to patients about everything“(…) giving clear instructions on what should be done really improves people’s behaviour”. (MD2)Reported by all professionalsProfessional interactions—networksProximity and easy communication channels between sectors (eg, health sector and municipalities) allow a close follow-up of the referred patients“We thus built much easier conversation channels, because in having partners on the field it was much easier for us to say ‘Look, pay attention to whom we referred to the programme’, for example, ‘5 patients. They are: this, this, and this and they have this and this pathology and this and this characteristic, so…’ (…) as we had the vast majority of patients accompanied by partners in the field, we knew what was happening with them (…) we knew if they were attending sessions, if they weren’t attending, if they had any problems, because we had this type of feedback with some regularity”. (MD5)Advantages for PA consultation patients provided by partners“We made a partnership in which people didn’t even have to pay anything to use the facilities”. (EP2)Partnerships provided having a detailed mapping of the community PA resources and increased the diversity of PA options“(…) because they have a deeper knowledge of all the potential of the community than we have (…) the practical details of all the [physical activity] resources that, in this case, our Municipality offered, right? In other words, it is one of the main strengths that I have to point out about the project”. (MD1)Intersectoral partnerships (eg, health sector and municipalities) contribute to sustainable action on PA promotion“I think it cannot be just the Municipality that promotes physical activity and the Municipality was available and is available to continue these types of projects. This needs to be done regularly, this promotion is also carried out through contact not only in the consultation office, with the white coat, but also outside the health centre. I think this turned out well. It must continue”. (EP1)Joint actions between different sectors and between different hierarchical levels of health sector were sometimes difficult to put in place and delayed implementation“At the beginning, it was a big mess… Because, first, that’s what I was explaining, the change of the partner enterprise took a long time, then they never responded, then there were those bureaucratic things… you have to sign a protocol and then I don’t know what more (…) we were only able to refer them [patients] to the community resources at a much later stage, almost at the end of last year… we were unable to do so earlier because of this delay in the protocol (…)”. (MD3)Reported by all professionalsProfessional interactions—team processes (constraints)Division of tasks between professionals and other team processes had to be refined in loci and it did not always work as planned“First, [the patient] signed up and would, supposedly, be called… He/she was being called first by the exercise professionals and only then they would come to me. That was how we were working… but we had to change the strategy, first they would come to me and then they would go to the exercise professionals, because otherwise my subsequent consultations would go down the drain, and they would come back to me in the final phase of the evaluations (…)”. (MD2)When all team members are aligned towards the same goals (in core team and extended team), things happen in a good way“It has to be a team project, of the entire health unit and when I say ‘team’, it means the whole team: doctors, clinical secretaries, nurses… because everyone participates in our health unit… therefore, from the recruitment of patients, from… all of this is important. The team has to be aligned in what the goals are and it has to be grounded in the project”. (MD5)Reported by all professionalsIncentives and resources—assistance tools and materials (constraints)It is important to give a written PA prescription to patients“It was very important, that I still have everything written down there, what the doctor gave me and the exercise professional. It was very important for me”. (P5)Professionals’ training was considered good from the theoretical point of viewProfessionals’ training was considered insufficient regarding local implementation and behaviour change science“(…) the training, when we went there, was more from the theoretical point of view of the plan, more from the theoretical point of view of the benefits of physical activity, diabetes itself, depression itself and not so much the implementation of the plan, and then yes, I think this initial training could be more about implementation… it was okay, it was the first time, one didn’t know how implementation would be, perhaps, but there would be more implementation training and on some specific data included in the manual”. (MD1)Implementation manuals were comprehensive and clear“The manual was very good… it was a little long, but it was clear, that is, we basically had to rely a lot on the manual”. (MD1)Devices/Technology would be helpful to monitor patients’ PA levels“So we missed having a way, if we had, for example, a step counter that we could give to everyone and have this feedback, for example. That’s it. Essentially, evaluation instruments”. (MD3)Reported by all professionalsIncentives and resources—information system (constraints)Clinical data were not accessible to exercise professionals in the software, which was time consuming and difficult to implement“(…) in fact, you may be prescribing an exercise that is not suitable because there is a lack of information”. (EP5)It would be more practical having an exercise prescription feature in the electronic medical softwareThe absence of an exercise prescription feature in the electronic medical software“Obviously, if you say to me ‘Oh, if the exercise prescription sheet could automatically come out of the [electronic medical software]?’ It would be excellent, of course. If the medical doctor could have a section of exercises that, knowing their patient, he could even make a few clicks and say ‘Look, I’d be happy if you do these exercises at home, half an hour, 3 times a week’ is better than nothing, isn’t it? Therefore, this would make life much easier for all family doctors, not only speaking about the consultation, but for all family doctors”. (MD6)The automatic referral through the electronic system has to be improved“What didn’t work well: first, the referrals… there were several error codes that didn’t allow to go forward, and, therefore, most colleagues were unable to refer”. (MD4)The PA consultation scheduling through the electronic system is not automatic for both professionals and is very time consuming“These are boring things, aren’t they? Because, I mean, every time I wanted to schedule a patient I had to copy paste the patient number, then schedule on the agenda of one professional, then log out, log in, then confirm the address… So, this had to be faster, as happens for a patient with diabetes, he has an appointment, I make an appointment and it is automatically sent to the nurse, I give it to the patient, the address is correct, I’m not worried… It’s much easier, not that. These are the small things that are heavy, that consume time, that for the future and have to be improved for processes’ optimization. (…) we were never able to the scheduling in the [electronic medical software] in an integrated way. That is, for example, if I scheduled the appointment for myself, I would never automatically schedule it for the exercise professional. This never worked, although we tried later. I think I even spoke to one or another person from the [informatic services of the Ministry of Health] about when we started this wasn’t working… in fact, we never managed it and I confess that we ended up giving up because we didn’t see it resolved. And this meant I had the exercise professional’s login and password, so I would go in and schedule an appointment, go out, log in with my credentials and schedule an appointment for myself, and this is time consuming”. (MD3)The PA consultation electronic feature was useful“(…) the software feature worked well in the planned consultation model, which was: first, medical consultation; then consultation with the physical activity professional for prescription and monitoring of the individualized prescription. In this model, I think it worked well”. (MD4)Reported by all professionalsCapacity for organisational change—(lack of) priority of necessary changeOpen mind posture and priority given by regional/local health authorities and municipalities were paramountClosed mind posture and insufficient priority given by regional/local health authorities and municipalities“Our health unit works well overall, I can put it that way. I’m suspicious, but it works very well in general and they are very dedicated to… innovation and different things and projects, so… (…) then proposed the project, everyone immediately said yes and that, that was it, it was good, the fact that it was a project pilot and our unit being one of the few… it was immediately well received. We then had the collaboration of a technical assistant, then, of a secretariat who would collaborate with us, that is… then they also welcomed the exercise professionals very well, the fact that they also went there, then they made a point of also talking to the IT department, creating the profiles and all these things, everything went well… that is, there were no obstacles, there was unity, that is… it’s like I was saying, my own supervisor excused me that morning to do that and didn’t raise any problems. I can’t say there were problems there”. (MD1)COVID-19 care was prioritised over all other healthcare interventions“(…) the pandemic reached out and it was one of the things that cost me the most, several of these people, especially in terms of depression… what happened to them, right? Because, quite honestly, I think the vast majority got worst because they needed more and we had there many cases to be successful, but that’s it. It was the conditions that came to us. Maybe we can pick them up again and pick up where we left off”. (EP2)Very frequent among all participantsIntervention factors—observable benefits of the interventionEffectiveness of the intervention in changing PA behaviour and having positive health outcomes was very valued“(…) we had a patient with diabetes who was taking a certain type of drug, which on average lowers 1% of A1C hemoglobin and he had stopped taking that drug, for economic reasons… I don’t know, he had stopped taking it. We, with the physical activity plan, managed to get him to have exactly the same values in A1C hemoglobin without the drug, that is, then we managed… we ended up reducing… there you go, exchanging the therapeutic plan, only with physical activity we achieved the same results without this drug. In other patients, we ended up maintaining control of A1C hemoglobin, maintaining… drastically reducing the units of insulin they had to take daily. In patients with depression, in this particular lady I was talking about, I remember that she was initially on medication and in the final phase she was no longer on medication, she had already stopped her antidepressant medication”. (MD1)Very frequent among all participantsPatient factors—motivationPatients’ motivation readiness at the referral positively influences everythingPatients’ lack of motivation readiness at the referral negatively influences everything“(…) one of the things that was very important was, right when we presented the project (to the health unit’s team), it was… and because the family doctor is the person who best knows the patients who were referred for consultation, it was an established fact that it had to be someone motivated to, and, therefore, right there, this makes your conversation and your motivational work much easier in consultation, because in fact… I’m not going to enter here into an area that isn’t even mine, it’s more yours, the motivation part, the different phases of the motivational cycle, but what happened was that family doctors sent people who were actually motivated to exercise”. (MD5)Patients’ motivation profile influences implementation and was associated with their underlying pathology by implementersPatients’ motivation profile influences implementation and may be associated with their underlying pathology by implementers“Patients with diabetes are much more resistant to adhering to an exercise programme, even going to the consultation. Patients with depression, let’s say, are patients who have been ill for a short time. In other words, they have had depression for a year, two or even months… and patients with diabetes, before they had diabetes, already had behaviours that led them to have diabetes. Therefore, this is a question of much deeper behavioural change than the question of depression. While in depression, it is easier, it can sometimes be more difficult to implement and the person is very depressed, but the fact of going to that consultation is because the person realizes that they need help and want to be helped. The issue of diabetes… I sometimes had patients, even my own that I referred to, and I knew that patients with diabetes didn’t want to go to the consultation, because they said they didn’t have time, there were always a thousand and one excuses, while patients with depression no, it is much easier to motivate them to attend the consultation and, after attending the consultation, adherence is much better in patients with depression than with diabetes”. (MD3)Patients felt motivated through the process because of the compromise created by the PA consultation and the health team“(…) and then my motivation was created by the team that, fortunately, has accompanied me at my health unit and that, in fact, created the motivation for me to continue and not give up. I thought that from a certain point onwards they were the ones who deserved this effort… ah… more than me. Because I think, in fact, from the moment that… this commitment exists and what led me, for example, to participate and be here today and now… ahhh… it also has to do… it has to do with… ahhh… my recognition for the effort and work that is developed for us. So, I think that in some way has to be awarded too”. (P6)Very frequent among professionalsIntervention factors—(lack of) feasibility or compatibilityLack of time and incompatibility with other assistance activities in the health unit were common“(…) I also only had EP3 support with me, once a week, at that time and therefore, when that time a week… and it is frequent in smaller Units like the ones where I work… I have to make an appointment of urgency, having to consult elsewhere and not being able to have that consultation, which then makes it very difficult to articulate things to give people that answer in a timely manner, and, therefore, that ‘once a week’, when it is removed, with the number of users we have and all that, it was already difficult to get it… sometimes, the 3-month appointment, it wasn’t at 3 months, it would be at 4 or 4 and a half months”. (MD4)Need to formally attribute more week hours to PA consultation implementation than those initially predicted“(…) those hours that were thought to be done every month are utopian for the process that has to be done”. (EP4)Intern medical doctors have more time flexibility and availability to implement the PA consultation“(…) another thing I remember was: we also had time for this, that is, I had the advantage that my supervisor ended up letting me go so I could do it and she would do the [family medicine] consultations… but I would do it too, sometimes… there were periods of urgency, which is part of our internship, when we are away for hospital emergencies. My colleague, who was also there, was also at the end of her internship, also after… she also had to leave, then there was also a period of vacation there, that is, there was also some unavailability on our part and also on the part of the exercise professionals, that is, those who were on vacation (…) there were also some difficulties here in terms of everyone’s availability, the whole team and there were quite a few of us (…)”. (MD1)Very frequent among professionalsIndividual health professional factors—motivationTo be well succeeded, PA consultation implementers have to be motivated and have interest in the area of PA interventions“(…) having fun doing this, feeling enthusiastic about doing this, because this is a lot related with enjoyment too, isn’t it? Like in many other things. Therefore, this very important factor of motivation on the days when the patient does not come, or on the days when the patient comes and has done absolutely nothing that we asked him to do, a little similar to what happens with other things… I think that’s the advice I give… is, in fact, to embrace the project and be focused and want to have fun with it too, because it’s important (…) I already had this idea, as soon as I started medical internship, in implementing a consultation… it would be exercise medicine, sports medicine to target a little more… the so-called special groups, risk groups, vulnerable groups (…)”. (MD5)Professionals’ proactivity regarding implementation is pivotal“(…) I am very stubborn and go knocking on doors, sending emails to the regional Public Health office, going there, annoying, annoying, annoying, to say ‘Attention, we want to have this here’”. (MD5)Very frequent among professionalsProfessional interactions—team communication (constraints)Team meetings between the medical doctor and the exercise professional to discuss cases should be predicted in the intervention modelTeam meetings between the medical doctor and the exercise professional to discuss cases were not formally predicted in the intervention model“(…) if necessary, we would sit down and have these informal team meetings that lasted around 1 hour, 1 hour and a half. In other words, these chronic people like… who have a chronic pathology as you were mentioning and very well, it can’t be resolved in three steps, it’s not when I am talking to the patient, that I can have the miraculous solution… And then, I remember clearly having a case with very peculiar situation there and at the end of the day, the patient left and I didn’t even know very well what to say… he left, at the end of the day, and I said to MD5 ‘hey, let’s talk…’ and we had a 40 minutes talking about the patient, establishing things and so on, then… after two or three days, I called and talked to him”. (EP4)It is important to create effective communication channels between the PA consultation team and the community PA programmes’ coordinatorsWhen effective communication channels between the PA consultation team and the community PA programmes’ coordinators are not created, the implementation is hindered“(…) that was one of the things that failed. In other words, we did at the beginning, and this was even written and signed by both parties in the contract they established, in the protocol, but it never happened. Even in email exchanges, whenever I referred a patient I said ‘I remind you that it will be important for us to get in touch for feedback about patients…’, but I never received any response. I remember at the time, when we met, I was introduced to two people who, theoretically, would be the ones who would give feedback, but we never received official contacts nor did we ever receive any type of feedback, it was only from patients, always”. (MD3)Very frequent among professionalsIncentives and resources—health facilities (constraints)Having dedicated medical offices to implement the PA consultation is paramountLack of dedicated medical offices to implement the PA consultation“The implementation challenges: more basic things like infrastructures, the lack of infrastructure… having an office for the consultation was a challenge, because… there it is, spaces are limited, we have very rich health care activity and, therefore, we were able to find… therefore, when we talk about patients’ accessibility to consultations, it was a challenge for us to be able to articulate it within the team… and when we talk about ‘the team’, from clinical secretaries, to the nursing staff, then medical doctor, the exercise professional, therefore, having infrastructures so that we could allocate the consultation there at that time”. (MD5)Very frequent among professionalsIncentives and resources—human resources (constraints)When a consultation is dependent on a restricted team, implementation will not succeed in case of professionals’ turnover“(…) in the long term, I see some barrier due to the fact that I will also finished my internship and then… it was one of the reasons why we stopped recruiting patients, because I would then have no one to leave to, let’s say, the continuity of the consultation… that is, to don’t leave the patients halfway, I preferred to follow the ongoing ones only and do not leave any patients at the middle of the process, because then I'll simply leave”. (MD1)Provision of exercise professionals by municipalities and universities was very facilitating, although future hiring by the health sector would be advantageousThe lack of provision of exercise professionals by the health sector may cause uncertainty of the implementation in the long term“(…) they made available an exercise professional for many more hours than in other places, that I know… didn’t made possible or professionals who were there voluntarily… and, here, my exercise professional doesn’t… he continues to be paid by the Municipality, but great part of his work is the pilot project”. (MD6) Lack of human resources “(…) extend the hours, at our cost, to be able to move the project forward (…) I lack human resources to give me support”. (MD4)Very frequent among professionalsIntervention factors—assessment proceduresAssessment questionnaires were too long and very time consuming“(…) questionnaires were very elaborate and patients left us thinking ‘well, when am I going to leave?’ and the consultation had not yet started… (…) about the issue of the questionnaires, I think it was a barrier and that, therefore, when the consultation will be implemented and does not require such a long questionnaire, it could be shorter and done more quickly”. (MD3)This table presents the most frequently reported implementation determinants, ranked by the number of participants identifying each one—determinants mentioned by all participants (either across the entire sample or within a specific subgroup, as specified in column 1) and ‘very frequent’ determinants (cited by over three-quarters of participants, either across the entire sample or within a specific subgroup, as specified in column 1). The three key facilitators identified were: high acceptability of the intervention characteristics; high diversity of community PA resources (eg, PA programmes) and implementers’ skills of interpersonal behaviour. Three key barriers were: lack of referrals to the consultation; lack of organisational priority for PA promotion; incompatibility of the consultation with existing healthcare demands.EP, exercise professional; MD, medical doctor; P, patient; PA, physical activity. Most frequent implementation determinants of the PA consultation This table presents the most frequently reported implementation determinants, ranked by the number of participants identifying each one—determinants mentioned by all participants (either across the entire sample or within a specific subgroup, as specified in column 1) and ‘very frequent’ determinants (cited by over three-quarters of participants, either across the entire sample or within a specific subgroup, as specified in column 1). The three key facilitators identified were: high acceptability of the intervention characteristics; high diversity of community PA resources (eg, PA programmes) and implementers’ skills of interpersonal behaviour. Three key barriers were: lack of referrals to the consultation; lack of organisational priority for PA promotion; incompatibility of the consultation with existing healthcare demands. EP, exercise professional; MD, medical doctor; P, patient; PA, physical activity. Capability-related determinants Health professionals’ skills of interpersonal behaviour Within psychological capability, health professionals’ skills of interpersonal behaviour were universally seen as facilitators among implementers. Positive interactions with patients during PA consultations, such as providing support, care, encouragement and clear instructions, improved patients’ motivation and engagement. For instance, an exercise professional noted: (…) knowing how to listen to people, it is very important to know how to listen to them and know how to empathize. (EP5) No determinants were identified for physical capability. Opportunity-related determinants Four determinants from the physical environment influencing intervention implementation and, thus, related to physical opportunity construct, were frequently reported by interviewees, all regarding incentives and resources. Assistance tools and materials Implementers unanimously highlighted the facilitating role of assistance tools and materials. Training focused on local procedures and behaviour change, comprehensive and clear manuals and devices for step counting were emphasised. Providing a written PA prescription to patients was seen as paramount for patient engagement. An exercise professional described: I made a sheet that is very similar to a medical doctor’s prescription, isn’t it? And therefore, I wrote here, I don’t know, walking or something… the dose, let’s say, you know? And these were small strategies that I noticed that… (…) these facilitating elements that we discover throughout the process are very interesting and can actually contribute… (…) I gave you this example of making the plan layout as a medical doctor’s prescription. So I think things actually went very well in many cases. (EP4) Information system The PA consultation feature in the electronic information system was deemed useful but faced constraints, such as patient referral errors and scheduling issues, leading to time-consuming procedures, complicated by data protection rules preventing access to clinical data for exercise professionals. A medical doctor suggested simplifying data sharing: (…) [the exercise professional] must type everything that’s in the document, which he receives on paper… that’s it, and… maybe it could be a little easier… I don’t know… just add a little button where, in fact, the patient accepts sharing the information (…) it goes to the exercise professional’s display. (MD5) An exercise prescription feature within electronic medical software was also proposed: I think that it would be perfect to have something identical to the [electronic medical software] with a set of suggestions for physical exercise and that we could also add what we have in the community, to meet people’s needs. (…) This would be interesting. (…). in an automated way and then it [the exercise prescription] either went to the patients’ cell phone and they consulted it there, or we printed it and delivered it directly to the person. (EP5) Health facilities Having adequate health facilities, specifically dedicated medical offices for PA consultations (a pre-requisite for the implementation) was crucial for implementation. Human resources Municipal provision of exercise professionals was valuable, but challenges included insufficient human resources to meet demand and reliance on a limited number of professionals. Participants suggested involving multiple professionals and advocated for political action to integrate exercise professionals into the National Health Service: (…) it doesn’t have to be centred on one person only. I think it makes sense for two-three people, depending on the size of the Health Centre Group and eventually to be able to have this consultation (…) (MD5) (…) efforts to have professionals hired by the National Health Service for the health centres (…) So, from a political point of view, I think this must be thought through (…). (EP1) Several determinants linked to social opportunity, reflecting contextual factors enabling or constraining implementation behaviours, were frequently reported. This encompasses social norms and cultural factors shaping attitudes and behaviours. Intervention characteristics Intervention characteristics emerged as the most frequently mentioned determinant, garnering consensus among all participants. Both implementers and recipients widely accepted the consultation format and characteristics, suggesting alignment with existing societal norms, beliefs and values. Key facilitators included multidisciplinarity, patient-centredness, consultation length and frequency, follow-up calls, comprehensive assessments and flexible prescription options (individual or community-based). Biannual follow-up consultations were suggested for enhancement: What needed to be improved was the so-called monitoring, perhaps every sixsix months (…) (P10) Participants recommended enhancing implementation sustainability by revising the criteria for medical doctors in the consultation team and implementing a mentee format. This would involve supervising new team members without a Sports Medicine background under a qualified doctor’s guidance. One participant suggested: We shouldn’t rely solely on one colleague… There should be a colleague with Sports Medicine training, but you can have another colleague, right? That is being supervised by them. (MD5) Referral strategy The referral strategy to the PA consultation was unanimously mentioned as a barrier. While patients responded positively to referrals from their family doctors, PA consultation medical staff reported a lack of referrals from colleagues within the health unit, limiting patients’ numbers significantly: (…) referrals did not exist. (MD2) Participants emphasised the need for implementing strategies to boost patients’ numbers, including diversifying referral sources beyond family doctors. For example, one exercise professional described a locally adopted approach: (…) we held a meeting with all the medical doctors in the unit, where I was introduced, where it was explained the whole project, the inclusion criteria, the exclusion criteria as well… and from then on, we gave them homework: within their patients, to identify those who met the criteria to be referred. We made available an Excel file, same shared file for everyone on the network, and they began adding names in there (…). (EP4) PA opportunities or resources for practice The variety of PA opportunities or resources, from individual to group activities, was emphasised as crucial to meet diverse patients’ needs and preferences. Implementers universally mentioned it and patients highly appreciated it. For example, two patients with different PA preferences noted: Being in a group setting is completely different, people make new friends and get to know each other… it’s different… For me, it’s more positive. (P17) I would participate again, because, well, I could do it at home. (P14) However, scheduling conflicts during working hours were cited as a common barrier to group PA programmes, limiting patient participation. Networks or partnerships Establishing networks between the health sector and community, particularly municipalities, emerged as a universal determinant among professionals interviewed, acting both as barrier and facilitator. It enhanced communication and enabled better patient follow-up and problem-solving. For example, one exercise professional noted: It was very interesting, because I was not just confined within four walls, I went out into the field to see how things were working and find the best solutions, so that the connections could really work. (EP5) Partnerships offered benefits like access to neighbourhood PA resources maps and reduced or no costs for referred patients. However, they occasionally complicated organisational procedures, causing delays and coordination issues between parties. Another exercise professional mentioned: What took the longest was this partnership with [the Municipality enterprise on physical activity], which still took a few months, and, when it was set, there were some obstacles because patients arrived there with the letter we gave them and whoever was there had no knowledge of the partnership. (…) There was some difficulty getting started, but then from the moment that happened, it went well. (EP2) Prioritisation of PA promotion Local and regional prioritisation of PA consultation implementation varied among organisations. According to professionals, some health units prioritised PA promotion, receiving strong support from management, while others struggled to gain traction. During the COVID-19 pandemic, efforts shifted to pandemic control, resulting in the suspension of PA consultations, due to safety concerns from health authorities for at-risk patients visiting health units. This raised concerns among implementers about potential setbacks and challenges in programme resumption. For example, one exercise professional expressed concerns: Well, we haven't been in the health unit since… since March 13th. Our contact is by telephone with Doctor [MD4]. We are discussing, (…) trying to figure out what to do, whether to keep going or wait. This kind of thing really gets to me, because we’re losing. We’ve lost. We are losing them [the patients referred to the PA consultation] (…) We’re losing them and it’s going to be tough. (EP3) Intervention factors Regarding intervention factors, two characteristics were frequently highlighted by all participants. First, the observed benefits of the intervention in improving patient health outcomes were frequently mentioned as a major facilitator. Second, concerns were raised by medical doctors and exercise professionals regarding the feasibility and compatibility of the intervention with routine care. To enhance compatibility with primary care demands, there was emphasis on formally allocating working hours to the PA consultation and addressing concerns about time constraints for all assistance activities. For instance, one medical doctor stated: The problem is everything that comes before that, isn’t it? All the needs that the health units are asked to respond to, to reach certain objectives (…). (MD2) This also underscores why intern medical doctors were considered more suitable for PA consultation implementation, given their flexible working hours and lack of a fixed patient list. Team processes PA consultation team processes were frequently mentioned as crucial, with positive team climate and alignment towards the same goals enhancing implementation. This alignment was potentiated when teams were well organised and everyone had a role, felt involved and was well-informed. However, deviations in initial consultation-related tasks and work distribution, requiring adjustments, posed barriers. Under team processes, communication processes were emphasised by all professionals at both community and intrateam levels. Establishing effective communication channels, like routine updates between the health team and community PA coordinators (facilitated by formal networks with municipalities), was considered a significant facilitator. Formal weekly case discussions within the health team were also suggested for better coordination. One exercise professional mentioned: (…) if something is to be implemented consistently, having at least two moments a week where the team meets would be important. (EP2) Assessment procedures Assessment procedures associated with the pilot study negatively impacted implementation. The assessment questionnaires were deemed excessively time-consuming and disruptive for patients, affecting intervention processes and team organisation: The questionnaire, taking an hour or more, was difficult to manage… even patients felt it was too long, becoming tired by the end of it… (EP5) Organisational regulations, rules or policies Although only a medical doctor cited a determinant within organisational regulations, rules, policies category, it would probably constitute a pivotal facilitator within the primary healthcare context. He mentioned the existence of contractual indicators for PA prescription in health units: If at a national level a contractual indicator for prescribing PA is implemented, it [the implementation] definitely increases. It’s obvious. (MD2) This participant briefly highlighted that the setting of delivery, namely the health region to which the health unit belongs, can also positively or negatively influence implementation. One exercise professional (EP3) also mentioned that the health setting, in general, felt somewhat unfamiliar compared with more informal contexts. Motivation-related determinants Motivation of receivers and implementers Patients’ and professionals’ motivations were very frequently mentioned regarding the sustainability and success of the implementation. Particularly, professionals’ proactivity regarding implementation and the referral of motivated patients were highlighted as pivotal. One exercise professional explained: People were not motivated and sometimes said ‘yes’ to the family doctor, but then missed the appointment. (EP5) Implementers’ cognitions and attitudes Although not the most frequently mentioned, professionals’ positive cognitions and attitudes towards PA and the intervention may have played a significant facilitating role among implementers. Valuing PA for its positive health benefits, practising PA, identifying with the intervention’s principles and viewing the PA consultation as a valuable intervention nationwide were motivational factors for medical doctors and exercise professionals to implement the intervention. Receivers’ expectations Patients’ expectations regarding the intervention were less frequently reported either, but they may have significantly contributed to patients’ commitment and retention. Positive expectations regarding the health benefits of PA and the perception that the PA consultation would provide additional assistance in improving quality of life were commonly mentioned facilitators by patients. However, a few were disappointed not to engage in PA during the consultation. The identified implementation barriers and facilitators were generally shared among medical doctors, exercise professionals and patients, indicating broad agreement. However, some barriers were perceived differently among subgroups. Only medical doctors reported team communication issues, suggesting they experienced more coordination challenges. Implementers (doctors and exercise professionals) cited feasibility, resource shortages and system constraints, but these factors were not noted by patients, implying they primarily affected service delivery rather than patient experience. Goal II: linking the intervention determinants to intervention functions and policies Table 3 specifies the corresponding functions of future implementation strategies and policies affecting their delivery, according to the COM-B components and BCW. According to the BCW framework, future implementation strategies aiming to improve professionals’ skills of interpersonal behaviour should focus on education, training or other forms of psychological capability enablement. Policies targeting communication/marketing, guidelines launch or regulation/legislation measures could be considered. Table 3Most frequent implementation determinants of the PA consultation, framed within the components of the BCWCOM-B model componentsMost frequent implementation determinants of the PA consultationIntervention functions, according to the BCWPolicy categories, according to the BCWCapability (individual’s psychological and physical capacity to engage in the activity concerned. It includes having the necessary knowledge and skills)Physical capability (physical capacity and ability to perform the behaviour)NANANAPsychological capability (capacity to engage in the necessary thought processes—comprehension, reasoning, etc)2. Individual health professional factors—skills of interpersonal behaviourEducationTrainingEnablementCommunication /MarketingGuidelinesRegulationLegislationService provisionFiscalEnvironmental /Social planningOpportunity (all the factors that lie outside the individual that make the behaviour possible or prompt it)Physical opportunity (afforded by the environment)5. Incentives and resources—assistance tools and materials (constraints)5. Incentives and resources—information system (constraints)5. Incentives and resources—health facilities (constraints)5. Incentives and resources—human resources (constraints)RestrictionEnvironmental restructuringEnablementGuidelinesRegulationLegislationFiscalEnvironmental /Social planningService provisionSocial opportunity (afforded by the cultural milieu that dictates the way that we think about things—eg, the words and concepts that make up our language)1. Intervention factors—observable benefits1. Intervention factors—intervention components, characteristics or content1. Intervention factors—referral strategy to the consultation1. Intervention factors—(lack of) feasibility or compatibility4. Intervention factors—assessment procedures4. Professional interactions—networks4. Professional interactions—team processes (constraints)4. Professional interactions—team communication (constraints)5. Incentives and resources—PA opportunities (constraints)—community programmes and individual prescriptions6. Capacity for organisational change—(lack of) priority of necessary changeRestrictionEnvironmental restructuringEnablementGuidelinesRegulationLegislationFiscalEnvironmental /Social planningService provisionMotivation (all those brain processes that energise and direct behaviour, not just goals and conscious decision-making. It includes habitual processes, emotional responding as well as analytical decision-making)Reflexive motivation (involving evaluations and plans)2. Individual health professional factors—motivation3. Patient factors—motivationEducationPersuasionIncentivisationCoercionCommunication /MarketingGuidelinesRegulationLegislationService provisionFiscalAutomatic motivation (involving emotions and impulses)2. Individual health professional factors—motivation3. Patient factors—motivationPersuasionIncentivisationCoercionEnvironmental restructuringModellingEnablementCommunication /MarketingGuidelinesRegulationLegislationService provisionFiscalEnvironmental /Social planningThis table maps implementation determinants in two ways: (1) across the components of the COM-B model and (2) across intervention functions and policy categories, using the BCW framework. This mapping helps to explore implementation strategies (based on functions and policy categories) to address determinants within each COM-B model component. Since opportunity-related implementation determinants were the most frequently mentioned, future implementation strategies should focus on restriction, environmental restructuring and enablement (functions). These strategies should reflect system-level policies related to guidelines, regulation, legislation, fiscal measures, environmental/social planning and service provision.BCW, Behaviour Change Wheel; COM-B, Capability, Opportunity, Motivation-Behaviour; NA, not applicable; PA, physical activity. Most frequent implementation determinants of the PA consultation, framed within the components of the BCW This table maps implementation determinants in two ways: (1) across the components of the COM-B model and (2) across intervention functions and policy categories, using the BCW framework. This mapping helps to explore implementation strategies (based on functions and policy categories) to address determinants within each COM-B model component. Since opportunity-related implementation determinants were the most frequently mentioned, future implementation strategies should focus on restriction, environmental restructuring and enablement (functions). These strategies should reflect system-level policies related to guidelines, regulation, legislation, fiscal measures, environmental/social planning and service provision. BCW, Behaviour Change Wheel; COM-B, Capability, Opportunity, Motivation-Behaviour; NA, not applicable; PA, physical activity. To address the above identified opportunity-related determinants in future implementation strategies, focus should be placed on environmental restructuring, whether at the health unit level or at national level. This can be complemented with restriction or enablement measures. Policies targeting environmental and social planning, service provision, fiscal measures or guidelines/legislation/regulation courses should be considered. To improve motivation-related determinants in future implementation strategies, focus should be on suitable education (whether pregraduate or postgraduate), persuasion, incentivisation, coercion, environmental restructuring, modelling or similar enablement methods. Policies could target all these categories. Specific measures will be further detailed in the following section. Discussion This study investigated barriers, facilitators and future strategies for implementing a PA prescription consultation in Portuguese primary care, using the TICD and BCW frameworks. The PA consultation was a pioneering intervention in the national context. While some research has examined similar initiatives, this is the first study to focus specifically on the Portuguese healthcare system. It is also the first to include perspectives from both implementers and patients. A key finding is that the most frequently reported barriers are primarily structural and opportunity-related, suggesting that system-level strategies are most appropriate to enhance future implementation (figure 2). Main findings of the study. C, Capability-related; M, Motivation-related; O, Opportunity-related. Top barriers included an ineffective referral process, a conflict with existing healthcare demands and lack of organisational prioritisation. Main facilitators pertained to high intervention acceptability, diverse community PA resources and strong interpersonal skills among implementers. Although most themes were commonly reported by all participants, challenges related to service delivery—such as issues with feasibility, resource shortages and constraints within the informatic system—primarily impacted service delivery more than patient experience. Barriers and facilitators spanned all seven TICD domains, highlighting its value as a comprehensive tool for designing and improving intervention implementation. Regarding intervention characteristics, in 52 studies Mino et al29 identified the impact of PA referral schemes’ components on PA levels, scheme uptake and adherence. In their review, screening, a person-centred approach and brief counselling were the only positive predictors of adherence. Integrating their findings with our multidimensional perspective can provide more comprehensive insights for optimising intervention design and implementation. At the resources’ level, offering diverse PA options and establishing formal protocols supported patient access to community-based services. Although some units experienced delays in setting up these protocols, partnerships with municipalities and universities were key facilitators, echoing findings from a recent systematic review.17 However, only one study specifically analysing PA prescription interventions applied a qualitative design. Using semi-structured group interviews, it explored general practitioners’ and nurses’ self-perceptions regarding the implementation of PA on prescription interventions in primary care centres.30 However, it did not evaluate the implementation of a real-world intervention (instead, it focused on general perspectives) nor the perspectives of patients and exercise professionals. Even so, participants pointed to a lack of knowledge among health staff about external PA resources available for referrals. The study emphasised the value of building community and municipal networks for improving patient access to PA resources (eg, through economic incentives linked to healthcare prescriptions) and for involving specialists in the care of patients with specific conditions.30 These findings align with recommendations from WHO4 and the European Commission,31 which advocate for promoting PA through intersectoral collaboration and community programmes. While the intervention’s multidisciplinarity, patient-centredness and follow-up calls were praised, concerns were raised about its integration into primary care routines. Staffing shortages and heavy workloads in the Portuguese National Health Service,32 mirroring global healthcare trends,33 often undermined plans to reduce doctors’ caseloads to accommodate PA consultations. This frequently led participants to identify feasibility and compatibility as barriers. Patients’ referral to the PA consultation was also challenging, increasing the burden on implementers. In some units, interns helped alleviate burden, but their high turnover after specialisation threatened sustainability. Previous systematic reviews similarly identified lack of time, misalignment with local routines and high workloads as barriers to implementation.17 34 35 This underscores the need to adapt interventions to contextual characteristics. Two interview-based studies evaluating PA on prescription models indicated that primary care staff and management generally agree the intervention is time-consuming and may constrain other healthcare activities.30 36 However, neither included patients, and one did not assess a directly experienced intervention, limiting the scope of comparison. At the individual level, despite literature highlighting implementers’ lack of knowledge and skills as significant barriers,17 34 35 these factors were infrequently cited. The requirement for implementers to possess specific qualifications and pre-implementation training likely reduced the prominence of this barrier. Still, some participants noted that behaviour change training could be strengthened, suggesting that certain knowledge gaps remain. A relevant topic to explore is why implementation barriers are mainly structural, whereas facilitators tend to emerge at individual and group levels. This points to a mismatch: even proactive intervention adopters and positive group dynamics can be hindered by unsupportive systems. Individual and system-level factors coexist and interact, highlighting the need for future implementation strategies to clearly address system policies, besides individual or team-based factors only. This aligns with the view of Chater and Loewenstein,37 which contrasts individual-focused interventions (i-frame) with system-focused policies (s-frame). They argue that overemphasis on individual interventions, without addressing system issues, has been ineffective and diverts attention from more impactful systemic changes. They emphasise the role of behavioural research in supporting system-level solutions, noting that s-frame policies can be more effective when informed by i-frame research.37 The European Union Policy Lab has also recently recommended integrating behavioural insights early in the policy cycle to better shape complex systems and enhance policy effectiveness.38 Using the BCW, we identified strategies to address common barriers, which typically emerged at system level. Environmental restructuring, regulation and service provision policies could target opportunity and automatic motivation-related barriers. Suggested actions include expanding consultation coverage across the Health Centre Group to serve a greater number of citizens living in the area, as well as diversifying referral strategies to enhance referrals to the consultation (eg, creating an online file for family doctors to identify patients, reminder materials of inclusion criteria). Allowing direct patient booking could also be considered to increase referrals from family doctors. Restructuring regulations/legislation could improve integration into routine care. This may involve introducing national clinical standards or orientations regarding PA promotion, allocating human resources, creating additional paid time slots for PA consultations in professionals’ schedules or providing institutional incentives linked to the attainment of PA indicators, under a value-based healthcare model. Such measures would enhance the intervention’s compatibility with routine tasks and strengthen its prioritisation within organisations. Improving software features (ie, sharing clinical data with exercise professionals, preset prescription form) and training would support implementation. Strengthening team support at service provision level, particularly in task management, would enhance team processes. Table 4 outlines additional strategies. Table 4Systematisation of the most frequent implementation barriers and facilitators of the PA consultation, with examples of how to intervene to inform future implementation strategiesImplementation determinants (facilitators and barriers)What worked well/features to maintainChanges neededBarrier summaryExamples on how to intervene, according to the BCW1. Intervention factors—intervention components, characteristics or contentPA consultation model was an overall facilitating factor1. Intervention factors—referral strategyPatients followed the referral of their family doctorCompromised patients’ reach: serious lack of patients’ referrals for the PA consultationIntervention level:Restructuring the referral strategy: diversify and identify effective referral strategies at local level—for example, in one health unit, the PA consultation team shared the PA consultation’s inclusion and exclusion criteria with all health unit’s medical doctors and shared an online Excel file for them to identify, within their patients those who met the criteria to be referred—they started putting names and it ended up being a big list.Restricting access to non-motivated patients: recruitment should focus only on patients interested in receiving the intervention.Environmental restructuring: constant reminders (meetings, papers, etc) directed to family doctors about the new PA consultation and respective inclusion criteria.Expanding target groups: after the end of the pilot project and when intervention’s scalability is intended.Regulation and/or legislation level:PA consultation serving a Health Centre Group (not one health unit, only): increase the number of referrers and eligible patients. Providing financial incentives, such as establishing contractual indicators for PA consultation implementation.5. Incentives and resources—physical activity opportunities (constraints)Physical activity opportunities diversity, whether community resources or other compatible with individual practice, was overall considered facilitating factors2. Individual health professional factors—skills of interpersonal behaviourProfessionals’ behaviour (whether genuine interest and attention concerning patients’ status, or proactivity regarding problem solving) was overall considered a facilitating factor4. Professional interactions—networksNetworks and partnerships created regarding the intervention implementation were overall considered a facilitating factor4. Professional interactions—team processes (constraints)Implementation team dynamics in which all team members were aligned and worked towards the same goal was considered a facilitatorImplementation procedures and team dynamics in loci did not always work as planned and needed to be refinedGuidelines level:Strengthen the guidelines towards intervention’s management tasks for each team element.Service provision level:Strengthen PA consultation team support in loci, regarding tasks management.5. Incentives and resources—assistance tools and materials (constraints)Manuals were considered facilitating factors. Written PA prescriptions were considered facilitating factorsTraining resources lacked more focus on local implementation issues and behavioural scienceService provision level:Restructuring the training provided to all intervention implementers: more depth on local implementation issues (eg, through experience sharing with experienced pairs, local visits from the national coordination team, etc) and more training time dedicated to behavioural techniques content.5. Incentives and resources—information system (constraints)The PA consultation module of the clinical medical software has very potential that could facilitate implementation even moreMany features of the PA consultation module of the clinical medical software did not work as intended or neededService provision level:Restructuring the software’s PA consultation module: patients’ biometrical data sharing with exercise professionals compatible with Protection Data Regulations (eg, consent given by the patient); improve automatic referral of patients to the consultation; improve automatic and simultaneous PA consultation scheduling for both professionals; create a time-saving exercise prescription feature.6. Capacity for organisational change—(lack of) priority of necessary changeAfter the arrival of the COVID-19 pandemic, the public health emergency management, along with the lack of priority given to other health interventions, constituted significant implementation barriersRegulation and/or legislation level:Implementing national clinical guidelines for PA promotion (eg, clinical standards; clinical orientations).Environmental restructuring: regulate and/or legislate in order to safeguard minimum health promotion and disease prevention services during public health emergencies.1. Intervention factors—observable benefitsPositive intervention results were overall facilitating factors3. Patient factors—motivationHaving motivated patients receiving the intervention was considered a facilitating factor1. Intervention factors—(lack of) feasibility or compatibilityThe intervention implementation was, in many areas, difficult to reconcile with the health unit’s and professionals’ usual tasksRegulation and/or legislation level:Restructuring the health unit vs intervention functioning regulation and/or legislation: formally create a time schedule for the PA consultation within the health unit’s and professionals’ agenda and reorganise professionals’ usual tasks in order to avoid conflicting processes.Implementing national clinical guidelines for PA promotion (eg, clinical standards; clinical orientations).2. Individual health professional factors—motivationHaving motivated professionals delivering the intervention was considered a facilitating factor4 Professional interactions—team communication (constraints)Team communication across all levels of the intervention (between PA consultation team elements, with other health unit professionals, with coordinators of the community PA resources to which patients are referred to, etc) was globally considered a facilitating factorAdditional time for team meetings was not planned beforehand in intervention characteristics and was invested at a personal costIntervention level:Restructuring the intervention characteristics/functioning: preview formal weekly moments for PA consultation’s team meetings in the professionals’ schedules.5. Incentives and resources—health facilities (constraints)The availability of dedicated office spaces in the health unit for the consultation was considered a facilitating factorIn some health units, having a suitable and dedicated place to implement the PA consultation was a challenge sometimesEnvironmental planning level:Reinforce communication with health unit candidates in implementing the PA consultation regarding the inclusion criteria of having available dedicated spaces.5. Incentives and resources—human resources (constraints)Having human resources with more weekly hours than the minimum allocated to the PA consultation implementation (eg, internship medical doctors; exercise professionals with more time allocated, full time or in higher number), one administrative employee dedicated to the PA consultation administrative procedures, exercise professionals hired by the National Health Service, other medical doctors involved within a supervising model were facilitators mentionedSmall PA consultation team, lack of human resources and their turnover were important implementation barriersGuidelines and/or legislation levels:Environmental restructuring: increase the minimum number of team members and/or define substitutes in case of absence; introduce a tutorial model of functioning; increase the minimum number of implementers’ dedicated hours for the PA consultation.5. Incentives and resources—assessment proceduresAssessment procedures, particularly the very time-consuming questionnaire application, were important implementation barriersEnvironmental planning level:For pilot study purposes, a reduction in questionnaires’ number of items/scales to the minimum essential would be welcomed. For a routine PA consultation implementation outside the pilot project, these assessment questionnaires will not be needed.Examples of emerging implementation strategies to address implementation barriers are provided in this table, using the inputs from table 3.BCW, Behaviour Change Wheel; PA, physical activity. Systematisation of the most frequent implementation barriers and facilitators of the PA consultation, with examples of how to intervene to inform future implementation strategies Examples of emerging implementation strategies to address implementation barriers are provided in this table, using the inputs from table 3. BCW, Behaviour Change Wheel; PA, physical activity. Studies investigating implementation strategies to enhance the implementation of PA consultations in primary care are lacking. Gustavsson et al39 tested the feasibility of six implementation strategies in the Swedish PA on prescription model delivered in primary care. Management support, educational lectures, local champions and external facilitation by a researcher were the most impactful. However, these strategies were insufficient to ensure the sustainability of the PA intervention. These findings, alongside ours, provide guidance in translating similar interventions into primary care. Strengths and limitations This research assessed implementation determinants based on how frequently participants reported them, which may not reflect the objective importance of each factor. It relied on subjective perceptions and individuals’ retrospective experiences, which could lead to recall and social desirability biases, potentially overlooking key determinants. Additionally, women represented only one-third of the sample and no patients were recruited from the Algarve Health Region. The videoconference and phone call format of the interviews may have also limited engagement compared with in-person interviews, potentially resulting in less detailed responses. A significant strength of the study is its practical applicability to real-world settings. The findings provide clear inputs on enhancing the implementation of PA prescription-based consultations in primary care, making them valuable to policymakers and adaptable to similar contexts. Additionally, the use of validated frameworks and models (TICD, COM-B and BCW) to collect and interpret data increases confidence in the results. These frameworks recognise multiple levels of influence and interconnectedness of various determinants, offering a holistic understanding of the system. Considering its individual components, as well as the relationships among them, contributes to a better understanding of the integrated whole system.40 The qualitative approach employed is also beneficial. Unlike relying solely on survey questionnaires, which could bias results by restricting determinants’ selection,40 the qualitative method provides a broader and more nuanced understanding. Using diverse data sources, such as interview transcripts and monitoring notes, enabled data triangulation, thereby reducing observer bias. Future research Incorporating methods such as Intervention Mapping, concept mapping, conjoint analysis and system dynamics modelling can enhance the systematic design, selection and tailoring of implementation strategies by addressing key contextual factors.41 Although it remains uncommon to test the mechanisms through which strategies exert their effects, understanding these causal pathways would help to clarify these complex implementation processes. This approach goes beyond the identification of barriers and facilitators, enabling the design of more effective, multilevel implementation strategies tailored to specific settings. Another important area for future research is the economic evaluation of implementation strategies to identify the most cost-effective options.41 Considering our results, investigating the cost-effectiveness of financial incentives, such as contractual indicators for PA consultations, could be particularly relevant. This would enable comparison between the required investment and the potential healthcare savings associated with increased PA levels among patients. Conclusion The PA consultation programme faced several implementation barriers, primarily structural and opportunity-related. Barriers included low referrals from family doctors, integration challenges with existing healthcare demands, lack of organisational prioritisation, insufficient human resources, time-consuming assessments, software limitations and issues in teamwork distribution. Using the BCW, system-level strategies were proposed to address these barriers, such as incentives based on PA indicators attainment, expanding consultation coverage and diversifying referral strategies, increasing human resources and improving the information system. These findings contribute to understanding the complexities of implementing PA consultations in primary care and provide valuable insights for enhancing similar programmes. Supplementary material