Physical activity (PA) reduces the risk for many noncommunicable, lifestyle-related diseases.1 Because of its health benefits, PA improves daily functioning, well-being, and quality of life while also lowering health care costs.2 In patients, increased levels of daily PA lead to improved health and fitness, which results in the maintenance of functional independence and improved quality of life.3 Besides the effects of PA on morbidity, PA is effective in preventing premature mortality in different patient groups, with benefits at least comparable with drug interventions in patients living with lifestyle-related chronic diseases (eg, coronary heart disease, stroke, heart failure, and prediabetes).4 The American College of Sports Medicine uses the paradigm “Exercise is Medicine” (E = M) in its global initiative to increase awareness of considering PA as a treatment option.
Initiatives for implementing E = M have mainly focused on the primary care setting,5,6 but it has been suggested that E = M should also be part of the hospital care system (secondary and tertiary care) in terms of treatment and prescription.7 Although the literature shows that E = M is effective,8,9 it is not yet routinely applied in the Netherlands. Many barriers hinder the implementation process. Among others, lack of clinicians’ time is a key barrier.10,11 There are several opportunities to tackle barriers to optimize the sustainable implementation of E = M in routine clinical care.12 For example, E = M can be applied in routine clinical care by embedding an E = M tool in electronic medical records (EMR). By generating a tailored PA prescription, such a tool has the potential to facilitate the implementation of E = M without requiring extensive knowledge of local PA facilities or major time investments of clinicians.12,13 Yet, implementing E = M together with the use of such new tools is a complicated process. Implementation strategies should be selected and tailored to address the contextual needs for introducing and implementing such tools.
In the Physicians Implement Exercise is Medicine (PIE = M) project,14 we implemented E = M in clinical care, tailored to specific clinical contexts, using the implementation mapping protocol. The set of PIE = M implementation strategies consisted of: educating clinicians, embedding an E = M support tool within the EMR, availability of a lifestyle coach within the department, an overview of local E = M referral options, and project support and coordination. In the current study, we tested the set of implementation strategies in 5 subsequent pilots within 2 different departments of a Dutch university medical center: the departments of rehabilitation medicine and orthopedics. In this learning process evaluation, we assessed the feasibility of the set of 5 PIE = M implementation strategies.
Methods
The PIE = M Project and the Design of the Pilots
Details of the PIE = M project have been described elsewhere.14 In summary, the project developed implementation strategies to support the clinicians’ prescription of E = M in routine clinical care. The implementation strategies were developed in response to the clinicians’ reported perceived barriers.11 Due to differences in department context, we developed an implementation protocol.15 This protocol provided guidance through the stages of implementation (ie, adoption, cocreation, prepare implementation, during implementation, and sustainability) and included implementation strategies for: (1) educating clinicians, (2) E = M tool in EMR,16 (3) lifestyle coach within the department, (4) overview of local E = M referral options, and (5) project support and coordination. In Table 1, we list the implementation strategies for each component. By implementation strategies, we mean a set of targeted, coherent activities that enable the implementation of E = M in a hospital setting, together with the implementation of a newly developed E = M tool.
PIE = M Implementation Strategies of Individually Tailored Prescription of PA in Clinical Care
| Part | Strategies | Activities and materials |
|---|---|---|
| Educating clinicians | Small-scale training sessions to increase the knowledge and skills of the target group by providing information, instruction, guidance, and feedback Using educational materials to provide instruction and guidance | Team information meeting: An explanation was given about the PIE = M project, and considerations to implement E = M during treatment were discussed with clinicians. Individual instruction about the pilot implementation: Clinicians were individually trained on using the PIE = M-specific tool as part of the EMR and on applying related conversation techniques Deployment of opinion leaders/leaders per department Information letter: Clinicians from the department selected to participate in the pilot received a letter explaining the study and the pilot An instruction card about applying the E = M tool and referral options and a reminder card to provide PA advice during the consultation were handed out Video instruction was developed to serve as a reference for the implementation, and additional information was provided about PA advice, registration in EMR, and referral options |
| E = M tool in EMR | Adjusting the professional role of the practitioner by making PA behavior part of a consultation Use information technology that supports patient care Prepare patients for contact with health care providers Provide patient education to influence patient behavior through self-management or lifestyle interventions under the guidance of a professional | Patient selection: Patients designated for a consultation with a participating clinician received a digital survey. Patients <18 y old and patients with severe comorbidity or psychological complaints were excluded by indication by the treating physician Survey: Before a consultation with the clinician, patients were asked to complete a short survey on their body height and weight, daily PA pattern, motivation to change their daily PA level, and their wish to discuss PA. The survey was linked to the EMR, so clinicians could access the results electronically during consultation PA advice: Within the EMR, personalized E = M advice was generated, indicating diagnosis-specific benefits of PA The clinician could discuss the results of the survey with the patient and printed the personalized advice PA referral: If deemed relevant, the clinician could directly refer patients to a lifestyle coach located within the department |
| Lifestyle coach within the department | Use of patient education targeting self-management or lifestyle interventions under the guidance of a professional Multidisciplinary collaboration to improve patient outcomes Providing integrated care to make lifestyle change part of usual care | Lifestyle coaches are health care professionals working in the outpatient clinic, closely collaborating with clinicians. They qualify for lifestyle counseling and give eligible patients one-off consultations aimed at knowledge transfer and, most importantly, regional referral options |
| Overview of local E = M referral options | Knowledge management to improve patient care through additional information Use of educational materials for health care providers to provide instruction and guidance Provide patient education to influence patient behavior through self-management or lifestyle interventions under the guidance of a professional | An overview for clinicians and lifestyle coaches with contact details about referral options was offered. Internally: (1) The lifestyle coach, directly available during the outpatient consultation; (2) “Sports Counter” within the hospital, where an exercise counselor provides information about exercise options in one’s living environment: On appointment. Externally: (1) Regular (allied) health care (physiotherapist or movement agog; (2) “Combined Lifestyle Intervention,” with referral via the general practitioner; (3) exercise consultant (exercise coach, community sports coach, walking group, coordinator adapted sports, and coordinator older adults) in the living environment of the patient, with referral via a PA broker Handout for patients: Contact details of the referral were provided on paper, and information about supporting websites and apps was provided |
| Project support and coordination: | Quality management of the implementation to improve performance in practice Provide personal communication (resources), a coordinator, peer contact, and advice Feedback through direct and indirect observation of care | Floor coordinator during consultations Individual instruction mailing before a pilot day, including: which patients had received the survey, their response status, the room number of the lifestyle coach, the available printers, and contact details of the floor coordinator for questions Helpdesk: The researcher provided remote support by email or telephone (Bouma) |
Abbreviations: E = M, Exercise is Medicine; EMR, electronic medical records; PA, physical activity; PIE = M, Physicians Implement Exercise is Medicine.
The implementation was divided into 5 pilots between October 2019 and October 2020, which were carried out by different clinicians, reaching out to different patient populations within the outpatient clinics of the Department of Orthopedics and Rehabilitation Medicine of University Medical Center Groningen (Supplementary Material S1 [available online]). The implementation strategies used within each subsequent pilot were slightly different as improvements were made based on the outcomes of the plan-do-study-act cycle of each pilot.17 Details of the changes made are described in Supplementary Material S1 (available online). A schematic presentation of the timeline is presented in Figure 1.
—Timeline of the 5 subsequent pilots.
Citation: Journal of Physical Activity and Health 21, 9; 10.1123/jpah.2023-0625
Two or 3 clinicians participated in each pilot round. Together with the management of the departments, specific consultation slots were selected. These slots included specific patient groups in which PA was considered to be related to the health condition. During these dedicated slots, the consultations would be different from usual care in that (1) patients with an appointment during the slot would be asked to complete a short survey before consultation, (2) the clinician would discuss the personalized active lifestyle advice generated by the E = M tool in the EMR with the patient, and (3) a lifestyle coach was available for a referral directly after the consultation a few doors down the hall of the clinician’s office. The lifestyle coaches were trained and certified professionals and provided single-tailored lifestyle advice.
The implementation of the pilots went differently from what we described in the study design14 and implementation strategies manuscripts.15 The pilots presented in this study all took place at University Medical Center Groningen in the departments of orthopedics and rehabilitation medicine. Due to the COVID-19 pandemic and associated visiting restrictions, the pilots planned in the department of rehabilitation medicine in Amsterdam University Medical Centers were canceled.
Participants
The PIE = M project was introduced at a department meeting by one of the researchers (Bouma), after which clinicians were invited to participate in one of the pilots on a voluntary basis. Informed consent was obtained from all participating clinicians before the start of data collection. The study was performed following the Declaration of Helsinki. Ethical approval for this study was waived by the medical ethical committee of the University Medical Center Groningen (METc UMCG2017/517).
Study Design
For this learning process evaluation, we employed a mixed methodologies approach. Data were derived from questionnaires completed by clinicians, semistructured in-depth interviews, logbooks, and field notes made by researchers during the study. When we had both quantitative and qualitative data on a specific topic available, we combined the results and used the qualitative results to put the quantitative results more into context.
A Conceptual Framework for the Learning Process Evaluation
The learning process evaluation was structured following the 5 constructs of the Reach, Effect, Adoption, Implementation, and Maintenance (RE-AIM) framework18 (Table 2—Operationalization RE-AIM).
Operationalization and Methods Used in the Learning Process Evaluation Following the RE-AIM Framework
| Items | ||||||
|---|---|---|---|---|---|---|
| Education | E = M tool in EMR | Lifestyle coach within dep. | Overview of local E = M referral options | Project support and coordination | PIE = M project coordinationa | |
| Reach Absolute number, proportion, and characteristics of participating patients | FQ, LB, FI, RF | FQ, LB, FI, RF | ||||
| Effectiveness Impact of the set of PIE = M implementation strategies on the perceived success | FQ, LB, FI, RF | FQ, LB, FI, RF | FQ, LB, FI, RF | FQ, LB, FI, RF | FQ, LB, FI, RF | FQ, LB, FI, RF |
| Adoption Absolute number, proportion, and representativeness of participating departments and clinicians | BQ, RF | BQ, RF | ||||
| Implementation Fidelity, adaptations to the PIE = M implementation strategies, and experiences with the implementation | FQ, LB, FI, RF | FQ, LB, FI, RF | FQ, LB, FI, RF | FQ, LB, FI, RF | FQ, LB, FI, RF | FQ, LB, FI, RF |
| Maintenance The extent to which the implementation of E = M has become part of routine care | FQ, FI | FQ, FI | FQ, FI | FQ, FI | ||
Abbreviations: BQ, baseline questionnaire for clinicians; E = M, Exercise is Medicine; EMR, electronic medical records; FI, follow-up interview with a clinician at the end of the pilot; FQ, follow-up pilot questionnaire for clinicians at the end of the pilot; LB, logbook usage of E = M tool by the clinician; PIE = M, Physicians Implement Exercise is Medicine; RE-AIM, Reach, Effect, Adoption, Implementation, and Maintenance; RF, researcher field notes.
aAlthough the project coordination was not an implementation strategy, it was considered essential for the PIE = M project and, therefore, considered as part of the process evaluation.
Questionnaires
Clinicians completed a questionnaire at baseline and after the pilot had ended. The participants would receive a secure link to the survey by email. The baseline questionnaire contained questions on personal characteristics, demographic information, work experience, personal lifestyle, and current patient PA referral behavior. An English translation of the used questionnaires is available in Supplementary Material S2 (available online). After the pilot had ended, clinicians were asked questions regarding their experience with the PIE = M implementation strategies and future PA referral behavior.
Semistructured Interviews
After finishing the pilot phase, all participating clinicians were approached for a semistructured interview (Supplementary Material S3 [available online]). During the interviews, clinicians were asked to reflect on the perceived impact of and satisfaction with the PIE = M implementation strategies and the extent to which E = M had become an integral part of routine care. The interview guide was specifically developed for this study and was guided by the RE-AIM framework.18 To ensure that clinicians felt free to speak their minds, the interviews were conducted by researchers who were not involved in the organization of the pilots (Nassau and Nauta). During interviews, a research intern was present to make notes. The interviews were conducted face to face or by video call and lasted approximately 30 minutes. All interviews were audio recorded and transcribed verbatim. Before analysis, a codebook was made according to the 5 themes of the RE-AIM framework: reach, effectiveness, adoption, implementation, and maintenance. The research team coded each transcript using a predefined codebook (Bouma, Nassau, and Nauta). Because of deductive coding, some codes were changed slightly throughout the coding process to correspond better with the perceived realities of the interviewees.
Logbook
Participating clinicians were asked to complete a structured logbook for each patient directly after the consultation. For each patient, the clinicians were asked, “Did you discuss the E = M advice?” “Did you counsel the patient regarding a more physically active lifestyle?” and if yes, “Where did you refer the patient to?”
Field Notes
During the pilot implementation period, the researcher (Bouma) kept track of events that may have affected the implementation as well as of changes in circumstances, implementation, and coordination. Field notes were used for interpretation of the results.
Results
Characteristics of Participating Departments and Clinicians
A total of 12 clinicians participated in the pilots (7 Department of Rehabilitation Medicine and 5 Department of Orthopedics). One clinician did not complete the baseline questionnaire. Details of the participating clinicians are presented in Table 3. Baseline outcomes showed that the participating clinicians reported having a relatively healthy lifestyle (on dietary behavior, body mass index, PA, and tobacco use). When asked on a scale from 1 (never) to 10 (always) about whether or not an active lifestyle had been discussed during consultation, clinicians scored a 5 (range 3–7) on average for both rehabilitation medicine and orthopedics. The average score for the question “Do you refer patients for lifestyle advice?” was relatively low in both the departments of rehabilitation medicine (2; range 0–6) and orthopedics (4; range 2–6).
Characteristics of Clinicians Before the Start of the PIE = M Pilot
| Department of Rehabilitation Medicine | Department of Orthopedics | ||
|---|---|---|---|
| Number of clinicians | 7 | 4a | |
| Physicians | N | 2 (out of 14b) | 2 (out of 11b) |
| Specialist registrar/residents | N | 5 (out of 6b) | 0 (out of 7b) |
| Physician assistants | N | 0 (out of 1b) | 2 (out of 5b) |
| Years of working experience | Mean (range) | 8 (1–23) | 8 (0–18) |
| Sex | % Women | 43 | 25 |
| Age | Mean (range) | 38 (26–58) | 45 (31–64) |
| BMI | Mean (95% CI) | 22.2 (18.2–25.2) | 23.5 (21.6–25.8) |
| Perception of own body weight | N (%) Healthy | 7 (100%) | 3 (75%) |
| Perception about own dietary behavior | N (%) Healthy | 7 (100%) | 4 (100%) |
| Importance of a healthy diet for clinician | N (%; very) important | 5 (71%) | 3 (75 %) |
| 150 min/wk MVPA or more | N (%) yes | 5 (71%) | 4 (100%) |
| Perception of own PA level | N (%; very) active | 4 (57%) | 2 (50%) |
| Importance of PA for clinician | N (%; very) important | 7 (100%) | 3 (75%) |
| Tobacco user | N (%) yes | 0 | 0 |
| Situation before pilot: | |||
| Discussion of an active lifestyle with patients? | 1–10 (never-always) | 5 (3–7) | 5 (3–7) |
| Referral of patients for lifestyle advice? | 1–10 (never-always) | 2 (0–6) | 4 (2–6) |
| Existing referral options | Physiotherapist internal, Physiotherapist external | Physiotherapist internal, Physiotherapist external, Movementologist | |
| N (%; fully) agree | N (%; fully) agree | ||
| As a clinician, I think it is important to discuss an active lifestyle with my patients | 6 (86%) | 3 (75%) | |
| I consider it my duty/responsibility to discuss an active lifestyle with my patients | 5 (71%) | 3 (75%) | |
| My colleagues think it is important that I discuss an active lifestyle with patients | 2 (29%) | 3 (75%) | |
| My manager thinks it is important that I discuss an active lifestyle with my patients | 4 (57%) | 3 (75%) | |
| I have sufficient knowledge to discuss an active lifestyle with my patients | 3 (43%) | 1 (25%) | |
| I have sufficient skills to discuss an active lifestyle with my patients | 2 (29%) | 1 (25%) | |
| I think discussing an active lifestyle is effective in positively influencing a patient’s lifestyle | 3 (43%) | 2 (50%) | |
| I am motivated to discuss an active lifestyle with my patients | 6 (86%) | 3 (75%) | |
| Discussing an active lifestyle with my patients is in line with the policy of my team | 6 (86%) | 3 (75%) | |
| I have enough time to discuss an active lifestyle with my patients | 0 | 1 (25%) | |
| There are sufficient referral options to lifestyle interventions where I can refer patients to | 0 | 0 |
Abbreviations: BMI, body mass index; MVPA, moderate to vigorous physical activity; PIE = M, Physicians Implement Exercise is Medicine.
aOne clinician did not complete the baseline questionnaire. bTotal number working within the department.
When looking at the perceptions on the implementation of E = M (ie, attitudes, social influence, and behavioral control), many clinicians reported having low perceived behavioral control over the implementation of E = M (ie, knowledge, skills, time, and referral options; Table 3).
Learning Process Evaluation
Because of the overlap between the PIE = M implementation strategies and the topics of the RE-AIM framework, it was decided to present the outcomes of the learning process evaluation grouped by the PIE = M implementation strategies that were described in the “Methods” section: (1) educating clinicians, (2) E = M tool in EMR, (3) lifestyle coach within the department, (4) overview of local E = M referral options, and (5) project support and coordination. Some of the topics reported on went beyond the scope of the PIE = M strategies.
1. Education
During the instruction, the clinicians received an instruction card for the EMR, referral options, and a reminder card. Some clinicians reported, specifically, the added value of the EMR instruction card. After an introduction, the E = M tool within EMR was not considered complicated to work with. Two clinicians remarked that the instruction was planned too long before the actual pilots started and that they could not remember exactly the user instructions of the E = M tool in the EMR. The clinicians rated the instruction video 7.7 (range 6–9; Table 4), on average, on a scale from 1 to 10:our attention is usually focused on, eh, on the not. For example, on how to not burden a foot. Because there is a wound. Eh, instead of having a focus on what that person would be capable of doing. (Clinician Department of Orthopedics)
That [video] was a nice addition because you could see how someone else did [discuss an active lifestyle]. (Clinician Department of Rehabilitation Medicine)
Satisfaction With the E = M Tool and Implementation Strategies Among Clinicians
| Implementation strategy | Department of Rehabilitation Medicine n = 7 | Department of Orthopedics n = 5 | Quote | ||
|---|---|---|---|---|---|
| Did the patient survey in the EMR help you to discuss an active lifestyle with your patients? (1–10) | E = M tool in EMR | Mean (range) | 7.4 (6–8) | 5.8 (4–7) | “I’ve [ ... ] always asked because no one had, well, maybe one or two people [ ... ], completed that survey in advance. [ ... ] Then I always explained, because it was such a [consultation hour with a lot of time pressure], [ ... ], uhm, discussed that this pilot was running and that [ ... ] there was, therefore, a possibility to speak to a lifestyle coach right away. And that went well, but I did not go through that survey with people and then fill it in.” Quote clinician Department of Rehabilitation Medicine |
| How user-friendly was the PA advice? (1–10) | E = M tool in EMR | Mean (range) | 6.3 (5–7) | 5.2 (2–7) | “Yes, I think that [the physical activity advice] contains the core. It shouldn’t contain more because [the components of the advice] are the most important topics to [start the conversation].” Quote clinician Department of Rehabilitation Medicine |
| How useful was the video “Introduction to PIE = M” (1–10) | Education | Mean (range) | 7.7 (7–8) | 7.8 (6–9)a | “[the clinician in the video] showed how you could [discuss the results of the physical activity advice]. Uhm ... It didn’t take 10 minutes. [the clinician] spoke a few sentences, the patient answered, and [the clinician] gave the paper [advice]. That was very clear, and I tried to use the same structure [during my counselling]” Quote clinician Department of Rehabilitation Medicine |
| To what extent has the opportunity to directly refer a patient to a lifestyle coach helped you in your referring process? (1–10) | Lifestyle coach within department | Mean (range) | 8.3 (7–9) | 8.6 (8–10) | “I think I wouldn’t know how to [refer patients] if I should have to do it myself. Because I would quickly refer a patient to a physiotherapist, I think. While a lifestyle coach may see other possibilities,” Quote clinician Department of Orthopedics |
| Did you use the map with suggestions for local PA options? | Overview of local referral options | N (% yes) | 5 (71%) | 1 (20%) | “[I advised] them that they could better discuss [the map with local exercise options] with such a lifestyle coach. Because I [ ... ] did not remember how that worked with those people in the region.” Quote clinician Department of Rehabilitation Medicine |
| How useful was the map with suggestions for local PA referral options? (1–10) | Overview of local referral options | Mean (range) | 6.4 (5–8) | -b |
Abbreviations: E = M, Exercise is Medicine; EMR, electronic medical records; PA, physical activity; PIE = M, Physicians Implement Exercise is Medicine.
a4 responses. b1 response.
2. E = M tool in EMR
Selection of Eligible Patients
Consultation slots of 210 patients (n = 123 rehabilitation medicine, n = 87 orthopedics) were selected to pilot test the implementation strategies. These 210 patients received a link to the online patient survey (Table 5). During the consultation, clinicians often saw patients who were either sufficiently physically active, too frail to have a physically active lifestyle, had an overuse injury, or could not afford the orthosis needed to participate in PA. One clinician even argued that the university medical center—with usually relatively complex patients—may not be the best place to implement E = M.
Use of the E = M Tool in the EMR and PA Advice Provided by Clinicians During Consultation
| Rehabilitation Medicine N (%) | Orthopedics N (%) | |
|---|---|---|
| Invited patientsa | 123 | 87 |
| Surveys completed before consultationa | 41 (33%) | 30 (34%) |
| Surveys completed during consultationa | 18 (15%) | 15 (17%) |
| Surveys not completeda | 64 (52%) | 42 (49%) |
| PA advice discussedb | ||
| Yes | 62% | 47% |
| No | 38% | 9% |
| Missing | — | 45% |
| If PA advice was discussed, did you refer the patient?b | ||
| Yes | 36 (64%) | 14 (54%) |
| No | 20 (36%) | 12 (46%) |
| Where did you refer to?b | ||
| Internal lifestyle coach | 34 (94%) | 13 (93%) |
| External regular care | 1 (3%) | — |
| External intermediary | 1 (3%) | — |
| Missing | — | 1 (7%) |
| Duration of the PA advice – mean minutes (range)c | 4.3 (1–12) | 5.8 (2–15) |
| Printed PA advice? ((almost) always) | 29% | 0% |
Abbreviations: E = M, Exercise is Medicine; EMR, electronic medical records; PA, physical activity.
aData source: electronic medical dossier. bData source: logbook completed by clinicians during consultation. cDuring clinicians’ consultation, not including the consultation with the lifestyle coach.
Because some people, yes, they just come for osteoarthritis complaints, but they are not very fat or very ..., yes, then I am not necessarily inclined to discuss it. (Clinician Department of Rehabilitation Medicine)
Collection of Patient Information
Clinicians indicated the shortness of the patient survey and the tailored PA advice to be favorable. However, during some of the pilots, the number of patients who completed the PIE = M questions before the consultation was low (∼33%, Table 5). When the patient did not complete the patient survey before consultation, the clinician completed it during consultation (∼15%, Table 5). Because that took time, the procedure was changed in later pilots so that the patient survey was completed by the lifestyle coach together with the patient. A perceived barrier was that the location of the patient survey results in the EMR was not intuitive, indicating that the user-friendliness of the E = M tool was not positive, with 6.3 out of 10 in rehabilitation medicine versus 5.4 in orthopedics (Table 5). Because that took time, the procedure was changed in later pilots so that the patient survey was completed by the lifestyle coach.
And none of them had any need. So uhm, not even to fill out the survey. And answers varied from, uhm, “I’m doing enough already” or “we walk so often in the hospital,” “we always get a request to participate in something” or “Yes, I don’t think it’s necessary.” (Clinician Department of Orthopedics)
Diagnosis
Some clinicians learned that people with a healthy body weight sometimes also needed PA advice. One clinician used the answer to the motivation question as a starting point for discussing a physically active lifestyle. In addition, some patients who did not want to discuss their lifestyle with the clinician appeared open to a conversation with a lifestyle coach.Now, it was the case with two or so of those patients that it naturally came up. And you have to bring it up with the others now. Because that has nothing to do with the complaint that people come for. (Clinician Department of Rehabilitation Medicine)
Personalized PA Advice
Clinicians varied the extent to which they used the automatically generated PA advice with their patients. According to the clinicians’ logbooks, the tailored PA advice was discussed in 40% to 60% of the consultations (Table 4). Yet, based on interview data, we believe that this number might be lower as clinicians often left the tailored PA advice to the lifestyle coach. Others reported frequently forgetting to discuss E = M or did not have the time to do so during the consultation.
In general, little time was spent during the consultation to discuss the PA advice, especially if the patient survey still had to be completed by the clinician. The average duration of the E = M conversation during consultation lasted 4.3 (average Department of Rehabilitation Medicine, range 1–12) to 5.8 minutes (average Department of Orthopedics, range 2–15; Table 5). We asked clinicians how they perceived the time they needed to invest in the E = M conversation. If any other matters needed to be discussed for the medical treatment plan, there was often little time left for lifestyle advice. Different clinicians suggested that, besides PA, lifestyle information should be expanded to other lifestyle domains for a more complete picture.Because what will people do with it, they come home, and it will be thrown away in the trash, I think. (Clinician Department of Rehabilitation Medicine)
PA Referral
If deemed relevant, patients could be directly referred to a lifestyle coach within the department for additional support toward a more active lifestyle or referred to a PA intervention in the patient’s vicinity. The percentage of patients referred to these services, including the lifestyle coach within the department, differed. In rehabilitation medicine, 40% were referred, of which 94% were referred to the internal lifestyle coach (Table 5). In orthopedics, 28% were referred, of which 95% were referred to the internal lifestyle coach. Other referral options were internal, “Sports Counter,” and external, to regular allied health care and to a lifestyle program that combined PA and dietary advice covered by Dutch health insurers and PA consultants (often employed by a municipality).
3. Lifestyle Coach Within the Department
All clinicians were very positive about the availability and functioning of the lifestyle coach within the department. One clinician even reported escorting patients in person to the lifestyle coach.
Clinicians perceived that the lifestyle coach’s presence saved them consultation time as the lifestyle coach completed the survey and discussed the PA advice. Also, the lifestyle coach was considered to have more knowledge on possibilities for a more physically active lifestyle and local referral options and to have better PA counseling skills. Clinicians mentioned that the quality of the lifestyle coach and free-of-charge lifestyle advice could facilitate the uptake by patients. The lifestyle coach was very visible to the clinicians, and this presence on the department floor served as a reminder for clinicians.
The fact that the lifestyle coach was on site did facilitate a PA consult for the patient and the clinicians. Some patients needed a little more persuasion to visit the lifestyle coach:I get the idea that the people that would benefit most [from an active lifestyle] are most difficult to convert. (Clinician Department of Rehabilitation Medicine)
In some cases, clinicians would have liked feedback on the effects of E = M counseling to get insight into the approach’s effectiveness. One clinician, for example, wondered if 1 active lifestyle consult would be sufficient to change a patient’s behavior.And they did want to, well, some [patients] didn’t think it was necessary to [go to] the lifestyle coach. But I would tell them; go see [the lifestyle coach] to support science, and do me this favor. Well, they would eventually go [see the lifestyle coach]. (Clinician Department of Rehabilitation Medicine)
4. Overview of Local E = M Referral Options
In the follow-up questionnaire, clinicians in rehabilitation medicine more often reported to have used the flyer with local PA consultants (71%) compared with clinicians in orthopedics (20%, Table 5). In the interviews, 1 clinician especially appreciated having an overview of the local PA consultants. In general, clinicians found the overview of local PA options useful. They said the overview was handed out to a few patients, for example, when a patient did not have time to visit the lifestyle coach directly after the consultation. Usually, discussing the options of the local PA consultants was left to the lifestyle coach by the clinicians, which was in line with the protocol.
5. Project Support and Coordination
yes, [the support staff] was really on top of things, that helped me a lot. Because otherwise I am sure that for half of the patients that I did refer now, I could not have referred [those patients] in an optimal manner. (Clinician Department of Rehabilitation Medicine)
6. PIE = M Project Coordination
A barrier mentioned with respect to the project was that the pilot was running for a limited number of selected consultation hours. As a result, clinicians did not develop a routine to discuss E = M with all their patients. Some clinicians told us that they still discussed E = M with some patients (after the pilot), especially patients with a complaint easily linked to a physically inactive lifestyle. One clinician was particularly motivated to continue the discussion of E = M and still referred the patient to a lifestyle coach through a general email address.
Some clinicians reported not needing a tool to screen patients who may have benefited from physically active lifestyle advice. They reported that after the pilot ended, they started giving patients more general advice toward a physically active lifestyle.
Almost all clinicians discussed their participation in the PIE = M pilot with colleagues. One clinician exchanged tips and tricks regarding the pilot with a colleague. The main reason for not discussing the pilot was that it was not very prominent in their thoughts. Some clinicians also reported discussing the pilot with the lifestyle coach in the department (Table 6).
Maintenance of E = M Referral
| Department of Rehabilitation Medicine | Department of Orthopedics | ||
|---|---|---|---|
| (n = 7) | (n = 5) | ||
| Do you still discuss a physically active lifestyle with your patients? never (1)–always (7) | Mean (range) | 2.7 (1–4) | 4.6 (3–7) |
| Do you still refer patients for a physically active lifestyle? never (1)–always (7) | Mean (range) | 2.6 (1–4) | 2.6 (1–4) |
| Would you still like to use the E = M tool in EMR during your consultations? | N yes (%) | 6 (86%) | 4 (80%) |
| Would you recommend participation in the pilot project to your colleagues? | N yes (%) | 6 (86%) | 4 (80%) |
| Would you recommend participation in the PIE = M project to other university medical centers? | N yes (%) | 6 (86%) | 5 (100%) |
| Would you recommend participation in the PIE = M project to peripheral hospitals? | N yes (%) | 5 (71%) | 5 (100%) |
Abbreviations: E = M, Exercise is Medicine; EMR, electronic medical records; PIE = M, Physicians Implement Exercise is Medicine.
Discussion
The PIE = M project was aimed at the implementation of E = M in 2 departments in a Dutch university medical center. In this learning process evaluation, which was structured and implemented in accordance with the RE-AIM framework, we aimed to monitor and evaluate the implementation of 5 consecutive pilots. In general, clinicians experienced E = M as positive. Three of the 5 main implementation strategies were particularly appreciated by clinicians: the education on E = M, the lifestyle coach within the department, and the project support. Although clinicians appreciated the E = M tool in the EMR, they reported some barriers that hampered the use of the E = M tool in practice. Also, the overview of local E = M providers was used infrequently by clinicians.
Increasing the knowledge and skills of clinicians through education about giving PA advice and referrals was considered a useful strategy. Also, a starting point for a conversation about PA was not always seen in patients. For 1 clinician, this project provided an incentive to discuss a physically active lifestyle. “I usually started with, you know. There, there is also a project going on here that concerns PA.” To this end, knowledge enhancement about E = M among clinicians is advised as an important strategy.12,15 More emphasis should be placed on the use of E = M within the education of health practitioners.
The presence of the lifestyle coach within the department to facilitate E = M referral seemed essential for the referral rate. Clinicians suggested not to have 1-time PA advice but a more comprehensive approach. Ideally, a clinician refers to a lifestyle broker, who has an intake with the patient using the E = M tool, after which the needs for change are assessed by the lifestyle broker, who also facilitates a personal and coordinated transfer to a lifestyle intervention within the patient’s living environment. This personal and coordinated transfer should preferably include a follow-up consultation to assess whether the transfer was successful. Such a strategy is currently being applied in the Lifestyle Front Office for Integrating lifestyle Medicine in the Treatment of Patients study.19
The number of actual referrals to the lifestyle coach was quite good (40% rehabilitation medicine; 28% orthopedics). However, for various reasons, referring all eligible patients to a lifestyle coach still appeared to be difficult. Lifestyle medicine is often inhibited by time constraints and competing priorities for the treating clinicians.12 Clinicians indicated that having various referral options as backup during a conversation about PA was very useful. To bypass clinicians’ perceived barriers, such as lack of time, knowledge, and skills, a dedicated lifestyle front office in hospital care may provide a solution in optimizing patient-centered lifestyle care and connecting with community-based lifestyle initiatives.
It turned out that adequate project support during implementation was essential when implementing E = M in medical departments. Yet, implementing E = M together with the use of such new tools is a complicated process. In practice, a lot of project support was needed to organize E = M in clinical care properly. The project support we provided was extensive to ensure that the implementation was successful but, as such, may be not realistic for routine care. Therefore, good, organized structures and processes and a point of contact within the hospital for E = M are essential. The implementation process should be designed as efficiently as possible. Tools such as an instructional video or an E = M tool can help to provide support.
An E = M tool integrated into EMR was useful in facilitating PA advice and referral, especially when selecting suitable patients. The operationalization of the E = M tool did have 2 barriers that may need further improvements: When using the E = M tool, patient information was conditional. The low number of completed patient surveys was hampering the continuity of the pilots. Strategies to increase patients’ responses must be scrutinized and implemented at the departmental level. The second factor reported by many of the clinicians was that using the E = M tool in the EMR should be quick and simple. This latter finding aligns with our previous findings on the requirements of an E = M tool.16
For specific patient groups, the PA advice proved to be very suitable. However, not all patients seemed eligible for PA advice due to their condition, underlying complaints, functioning, motivation, capabilities, or vulnerability. To further support clinicians in discussing PA with their patients, more guidance toward specific PA opportunities for different patient groups would be helpful. This information should include health condition-specific evidence for E = M prescription and more guidance for opportunities to tailor PA advice toward people with different health conditions.11
Clinicians suggested that, besides PA, lifestyle information should be expanded to other lifestyle domains for a more complete picture. This aligns with other research13 wherein identifying several lifestyle domains is advocated for personalized lifestyle advice. Providing help should connect with observations during consultation. For example, a clinician said: “Some factors make the conversation more likely, for example, if a patient is overweight or smells like smoke.” If attention is paid not only to PA but also to lifestyle in a broad sense, the advice will have a greater impact on the health of the patient.
With the learning process evaluation based on the RE-AIM framework, we can draw conclusions about the current implementation. Reach: The involvement of the clinicians in the departments was high. We involved a large number of patients in the project, of whom a substantial number were referred to a lifestyle coach for advice or external referral. Effectiveness: The impact of the different implementation strategies varied between the strategies. The highest impact was achieved with education of the clinician, the presence of the lifestyle coach within the department, and project support. This supported the clinicians in their behavioral control, previously indicated as low, by providing knowledge, by increasing their conversational and referral skills, and by the possibility of having assistance from the lifestyle coach. Adoption: The departments that were already active on “exercise” have now participated in the project. Adoption in these departments was reasonably good. But these departments may not be representative of all other hospital departments and clinicians. More effort is needed to implement E = M broadly across the entire hospital. Implementation: The importance of implementation is clearly endorsed, and clinicians are motivated by the topic. We have learned that the implementation must be made as easy as possible for clinicians, taking little time, being easy to understand, and requiring as few additional actions as possible. Carrying out E = M must fit into the clinician’s existing work process. Maintenance: The importance of E = M is strongly felt by many clinicians, but E = M has not yet been implemented well enough in usual clinical care to ensure that it will now be maintained. The implementation of E = M will only be sustained if the positive outcomes outweigh the small investment for implementation. Limitations: The participating clinical departments (Orthopedics and Rehabilitation Medicine) may have had a stronger focus on stimulating PA in patients than other clinical departments of a hospital would do. Because participants from departments voluntarily registered for the pilots, more motivated clinicians likely participated, resulting in a smoother implementation than in a real-world setting. Moreover, clinicians had relatively healthy lifestyles themselves. For that reason, they may be more inclined to talk about PA with patients and encourage them to adopt a healthy lifestyle. On the other hand, starting with positively minded departments and clinicians may promote implementation.
In this study, we describe that 5 main implementation strategies have been employed. It should be noted that the involvement in the project and the employed research methods (eg, using logbooks) may also facilitate implementation.
The COVID-19 pandemic forced us to deviate from our original plans presented elsewhere.14 After the restart, more consultations were online, which made discussing PA and referrals more difficult. We noticed that clinicians were deprioritizing the project due to the pandemic. As a result, fewer participants participated in the project, which hampered implementation.
The patients’ own experiences are not included in this study. If we want to continue with this integration in health care in the future, it is important that we also include the perspective of the patient.
Conclusions
In the current study, we determined the feasibility of a set of strategies for implementing E = M in clinicians’ routine work processes. We showed that specific implementation strategies, such as education on E = M, a lifestyle coach within the department, and project support and coordination, are effective in facilitating the implementation of E = M. The working process should be carried out very efficiently because of several perceived barriers. E = M tools, preferably embedded in the EMR, could assist clinicians with patient selection and easy referral options.
Acknowledgments
Author Contributions: Bouma and Nauta shared first authorship. We would like to thank the following interns for their support during the various phases of the PIE = M project: Yvon Douma, Anouk Driessen, Äaron Spapens, Nanick van der Wal, and Kim Wolffenbuttel. And we would like to extend our gratitude toward all participating clinicians, department managers, lifestyle coaches, and patients from orthopedics and rehabilitation medicine at University Medical Center Groningen and Rehabilitation Medicine at Amsterdam University Medical Center. The authors submitted this paper on behalf of the PIE = M consortium. The members of the PIE = M who were not included in the author list are: The Lifelines Cohort Study, the Netherlands: W.J.R. Bossers. University of Groningen, University Medical Center Groningen, Department of Cardiology, Groningen, The Netherlands: J. Brüggeman. University of Groningen, University Medical Center Groningen, Department of Internal Medicine, Groningen, The Netherlands: G.J. Navis, S. Scholtens. University of Groningen, University Medical Center Groningen, Genomics Coordination Center, Groningen, The Netherlands: M.A. Swertz. University of Groningen, University Medical Center Groningen, Center for Human Movement Sciences, Groningen, The Netherlands: K.J. van der Velde. Funding: This work was supported by a grant (grant 546001002) from ZonMW.
References
- 1.↑
Ekelund U, Steene-Johannessen J, Brown WJ, et al. Does physical activity attenuate, or even eliminate, the detrimental association of sitting time with mortality? A harmonised meta-analysis of data from more than 1 million men and women. Lancet. 2016;388(10051):1302–1310. doi:
- 3.↑
Bouchard C, Shephard RJ. Physical activity, fitness, and health: the model and key concepts. In: Bouchard C, Shephard RJ, Stephens T, eds. Physical activity, fitness, and health: international proceedings and consensus statement. Human Kinetics Publishers; 1994:77–88.
- 4.↑
Naci H, Ioannidis JP. Comparative effectiveness of exercise and drug interventions on mortality outcomes: metaepidemiological study. BMJ. 2013;347:f5577. doi:
- 5.↑
AuYoung M, Linke SE, Pagoto S, et al. Integrating physical activity in primary care practice. Am J Med. 2016;129(10):1022–1029. doi:
- 6.↑
Shuval K, Leonard T, Drope J, et al. Physical activity counseling in primary care: insights from public health and behavioral economics. CA Cancer J Clin. 2017;67(3):233–244. doi:
- 7.↑
Sallis RE, Matuszak JM, Baggish AL, et al. Call to action on making physical activity assessment and prescription a medical standard of care. Curr Sports Med Rep. 2016;15(3):207–214. doi:
- 8.↑
Lobelo F, Rohm Young D, et al. Routine assessment and promotion of physical activity in healthcare settings: a scientific statement from the American Heart Association. Circulation. 2018;137(18):e495–e522. doi:
- 9.↑
Schmitz KH, Campbell AM, Stuiver MM, et al. Exercise is medicine in oncology: engaging clinicians to help patients move through cancer. CA Cancer J Clin. 2019;69(6):468–484. doi:
- 10.↑
Clark R, McArthur C, Papaioannou A, et al. “I do not have time. Is there a handout I can use?”: combining physicians’ needs and behavior change theory to put physical activity evidence into practice. Osteoporos Int. 2017;28(6):1953–1963. doi:
- 11.↑
Nauta J, van Nassau F, Bouma AJ, et al. Facilitators and barriers for the implementation of exercise are medicine in routine clinical care in Dutch university medical centres: a mixed methodology study on clinicians’ perceptions. BMJ Open. 2022;12(3):e052920. doi:
- 12.↑
Bowen PG, Mankowski RT, Harper SA, Buford TW. Exercise is medicine as a vital sign: challenges and opportunities. Transl J Am Coll Sports Med. 2019;4(1):1.
- 13.↑
Stout NL, Brown JC, Schwartz AL, et al. An exercise oncology clinical pathway: screening and referral for personalized interventions. Cancer. 2020;126(12):2750–2758. doi:
- 14.↑
Krops LA, Bouma AJ, Van Nassau F, et al. Implementing individually tailored prescription of physical activity in routine clinical care: protocol of the Physicians Implement Exercise = Medicine (PIE = M) Development and Implementation Project. JMIR Res Protoc. 2020;9(11):e19397. doi:
- 15.↑
Nassau FN, Bouma AJ, Krops LA, et al. Stepwise development of an implementation protocol to support prescription of Exercise = Medicine by clinicians: application of the Implementation Mapping protocol. Under review.
- 16.↑
Bouma A, van Nassau F, Nauta J, et al. Implementing Exercise = Medicine in routine clinical care; needs for an online tool and key decisions for implementation of Exercise = Medicine within two Dutch academic hospitals. BMC Med Inform Decis Mak. 2022;22(1):250. doi:
- 17.↑
Edelenbos J, Van Buuren A. The learning evaluation: a theoretical and empirical exploration. Eval Rev. 2005;29(6):591–612. doi:
- 18.↑
Glasgow RE, Vogt TM, Boles SM. Evaluating the public health impact of health promotion interventions: the RE-AIM framework. Am J Public Health. 1999;89(9):1322–1327. doi:
- 19.↑
van Dijk ML, Te Loo LM, Vrijsen J, et al. LOFIT (Lifestyle front Office For Integrating lifestyle medicine in the Treatment of patients): a novel care model towards community-based options for lifestyle change—study protocol. Trials. 2023;24(1):1–21. doi:
