Over the years, many research projects related to technology implementation in the context of chronic diseases have been developed in primary care. One of the main effects is to contribute to a favorable lifestyle medicine change, which may have positive repercussions on patient health and the management of chronic diseases [1-4].

Recently, a scoping review by Clarkson et al [5] identified the increased use of digital tools, in combination with human support, to help people with long-term conditions and to maintain physical activity (PA). This review shows that web-based digital tools continue to predominate with the more recent emergence of gamification, applications, and virtual environments. However, most participants were from younger age groups, the use and description of the theory in the development of the tools were limited, and most studies highlighted the need for human engagement to support their use [5]. The lack of digital tools for multimorbid long-term conditions, longer-term follow-ups, understanding participants’ experiences, and informing future questions about the effectiveness were the obvious gaps [5].

Another scoping review, conducted by Motahari-Nezhad et al [6], revealed that clinical evidence concerning digital biomarkers has been systematically reviewed across a wide range of study populations, interventions, digital devices, and sensor technologies with the dominance of PA and cardiac monitors. To understand the clinical value of digital biomarkers, the strength and quality of the evidence on their health consequences need to be systematically evaluated.

Indeed, the use of digital technology to help patients with chronic diseases such as type 2 diabetes is an emerging field of research. There is a variety of equipment available to patients, including consumer activity trackers, pedometers, smartphone apps, and blood glucose monitors. The advantages of these technologies are that they can allow health care professionals to remotely monitor patients and reduce the need for patients to regularly attend clinics [7]. Recently, a mixed methods study by Hodgson et al [4] explored the use of an activity tracker for 4 weeks to support an active lifestyle in adults diagnosed with type 2 diabetes. Overall, the results demonstrated that fitness trackers could support an active lifestyle in adults with type 2 diabetes, but more detailed discussions with health care professionals could identify methods of integrating activity trackers into patient care [4]. Additionally, a multicenter prospective observational study (set up for 7 weeks) has evaluated the feasibility of using a nonmedicalized device to monitor the lifestyle of elderly patients with type 2 diabetes in a primary care setting [8]. Researchers observed a high level of acceptance of portable devices based on the impressions of patients and health care professionals [8]. Nevertheless, the authors suggest further studies to assess devices’ acceptability for longer periods [8].

Additionally, it is important to point out that technical failures were faced in many studies, such as log-in difficulties [3], connectivity errors [3,9], a lack of reliability and validity [3,9,10], as well as some patients feeling overwhelmed by the technical complexities of the activity tracker and its software [4,10]. Patient motivation dropped when such complications occurred [3]. The limited resources, especially in terms of staff [9,11] and money [9], were also reported as barriers to implementation.

A hybrid type 1 trial aims to determine the effectiveness of clinical intervention and to better understand the implementation context [12]. It is useful to explore if a clinical intervention works in a specific context and to gather potential barriers or facilitators to an intervention’s widespread implementation. Based on this design, a randomized pilot trial was conducted to evaluate the impact of an activity tracker on PA and cardiometabolic variables in a real-life settings among patients with type 2 diabetes [13,14]. Alongside this randomized pilot trial, we tested its effects on relevant outcomes while observing and gathering information from patients and the health care team with a strengths, weaknesses, opportunities, and threats (SWOT) analysis [15]. SWOT analysis has been used to determine the components of successful technology implementation in primary care settings, and its objective is to use the knowledge an organization has about its environment to formulate its strategy accordingly [15]. To the best of our knowledge, no study has simultaneously studied the feasibility and implementation among patients with type 2 diabetes using an activity tracker to increase motivation for PA in primary care with a SWOT analysis.

The aim of this study is to carry out a SWOT analysis to assess the acceptance of this novel technology by (1) assessing satisfaction among patients with type 2 diabetes using an activity tracker to increase motivation for PA and (2) exploring the health care team’s perception of its implementation in a primary care setting.

We conducted a hybrid type 1 study for 3 months in an academic primary health center in Quebec City, Quebec, Canada. The study consisted of 2 stages. The first stage was a pilot randomized controlled trial to test a clinical intervention, and the second stage was to gather information on its implementation potential with patients’ and the team’s feedback using a SWOT analysis.

The first component consisted of a pilot randomized controlled trial of 30 patients with 2 groups of 15 patients each to evaluate the impact of an activity tracker on PA and cardiometabolic variables in a real-life context among patients with type 2 diabetes. The complete study protocol was published on ClinicalTrials.gov (NCT03709966) [13]. The results of this study are published elsewhere [14]. The control group received routine follow-up, which included a PA promotion intervention supported by a kinesiologist. The intervention group is routinely followed up with the addition of an activity tracker (Fitbit Charge HR, Fitbit Inc) worn during the study. Participants in the intervention group were also provided with a tablet (iPad, Apple Inc) and with an application linked to the activity tracker (the Fitbit app). Cardiometabolic risk variables, PA motivation, and PA were assessed at the baseline and at the end of the trial.

The second component of the study consisted of gathering information on the delivery and potential of implementation. Feedback from patients and the health care team was gathered to perform a SWOT analysis. A SWOT analysis was particularly relevant in the context of this project in order to identify factors that could influence the replication of this intervention in a similar context. The SWOT analysis considers factors that are internal (strengths and weaknesses) and external (opportunities and threats) to the intervention. Thus, it can inform on both the design of an intervention and the strategic planning of its implementation [15].

Satisfaction and acceptability of the activity tracker as well as the application were measured by a questionnaire inspired by the Technology Acceptance Model [16] and the System Usability Scale [17,18]. The 15 participants in the intervention group were invited at the end of the study to fill out the 10-question questionnaire on their satisfaction with the device and the support provided by the research team, their opinion about the information the device is providing, the impact on their lives, and their comments on the study. The questionnaire was in French and contained both quantitative and qualitative elements. The questionnaire is available in Multimedia Appendix 1.

Research team members and health care professionals who were engaged in this project were also consulted. We used LimeSurvey, a web-based survey tool, to collect the team’s feedback [19]. In total, 12 participants and team members were invited to fill out an original questionnaire based on the SWOT analysis elements [15]. It contained 11 questions about their identification, the study’s strengths and weaknesses, the achievement of its objectives, the barriers faced, the facilitating factors, the integration of an activity tracker in primary care, and possible improvements. The questionnaire was in French and contained both quantitative and qualitative elements. A link to fill out the questionnaire was sent via email, and a reminder was also emailed a month later. The questionnaire is available in Multimedia Appendix 2.

The authors are accountable for all aspects of the work. The trial was conducted in accordance with the Declaration of Helsinki (as revised in 2013) [20]. The study was approved by the institutional ethics board of the Centre intégré universitaire en santé et services sociaux de la Capitale-Nationale (No: 2017-2018-07), and informed consent was obtained from all the participants. Data were deidentified to preserve confidentiality. Furthermore, the data from the activity tracker are subject to the Fitbit privacy policy [21]. No financial compensation was given, but parking tickets for the clinic were provided.

Quantitative variables from selected answers (satisfaction and acceptability of the activity tracker’s questionnaire) were reported in frequency tables. Qualitative variables from open questions (the implementation questionnaire) were compiled in a file, and key concepts were identified by hand by the first author (CP). The first author then manually performed an inductive thematic analysis [22] to regroup key concepts as they emerged from data from both questionnaires, which were validated separately by 2 other coauthors (CR and MPG). The themes were then synthesized in a matrix and ranked according to apparition frequency and global importance. The information gathered was triangulated using the SWOT model (Opportunities-Strengths [O-S], Opportunities-Weaknesses [O-W], Threats-Strengths [T-S], and Threats-Weaknesses [T-W]) to propose recommendations that were then approved by the research team. For quantitative variables, normality was assessed using the Kolmogorov-Smirnov test. Variables following a normal distribution were expressed as mean (SD) values; otherwise, median (IQR) values were used. Missing data were excluded. The analyses were performed with the software SAS (version 9.4; SAS Institute).

The first stage of the study, the randomized pilot trial, showed that PA assessed by questionnaire increased in the group with a PA intervention supported by a kinesiologist (the control) and in the group with an activity tracker in addition to the PA intervention (the intervention). High-density lipoprotein cholesterol increased in the intervention group and decreased in the control group (P=.01). Glycated hemoglobin tended to decrease in both groups (P=.08). The full results are published elsewhere [14].

The improvements suggested by the research team and the participants were classified according to the O-S, O-W, T-S, and T-W strategies (Textbox 1). The main points for O-S were (1) integration of activity tracker data in the electronic medical record, (2) organization of academic health centers, and (3) follow-up on a longer period. The main points for O-W were (1) a larger study with more participants; (2) extraction of all the activity tracker data; (3) collection of data in pre-post intervention, clarification of objectives; and (4) formation of groups of motivation. The main points for T-S were (1) full-time team dedicated and available to the clinical intervention, (2) addition of more academic health centers, and (3) limit the number of professionals involved. The main points for T-W were (1) full-time research coordinator, (2) scientific counselor, (3) training on how to use the device, (4) team training with the technological tool, and (5) logging access given by the team.

Overall, the team members felt that the implementation in primary care was feasible:

Another team member mentioned that “it [activity tracker] brings objective data on PA and could improve physician-patient communication” [Team member 3]. One team member felt some people might benefit more from it: “I think that this device is interesting mostly in cases of instability or huge variability in results/medical analyses of patients. The device allows us to see fluctuations and potentially establish certain tendencies” [Team member 5].

This is a 3-month pilot study conducted in Quebec, Canada, that aimed to assess the satisfaction of patients with type 2 diabetes who use an activity tracker to increase motivation for PA, as well as to investigate the patients’ and team's perception of the implementation of this technology in a primary care facility. According to this research, patients with type 2 diabetes were satisfied with their activity tracker, which was used to increase PA motivation. Members of the health care and research teams agreed that this novel technology tool can be used in primary care, although there are still obstacles to its frequent use in clinical practice.

Improvements included the integration of relevant activity tracker data in the electronic medical record, involving more family medicine units, repeating the study with more participants, and hiring a full-time research coordinator. Most of the team members perceived that an activity tracker could be integrated into the follow-up of patients in primary care. Overall, participants appreciated the activity tracker, as did the team, but as shown by the SWOT analysis, there are some challenges. The results obtained are consistent with the preliminary literature regarding the implementation of activity trackers in primary care. Reed et al [9] reported an increase in PA among inactive rural adults when implementing a 12-week Fitbit-based intervention and technological difficulties, while the nursing staff cited human resources and money as barriers. Similar work has been completed among both patients who had diabetes earlier [36] and those with diabetes and other cardiometabolic conditions reported in a recent systematic review and meta-analysis [37].

We believe that to facilitate digital health research in primary care, it is essential to provide a structure dedicated to supporting clinical research using digital technologies, which requires an adequate organization of academic health centers, substantive funding, and human resources. The clinical research coordinator supports and oversees the daily activities and plays a critical role in the conduct of the clinical trials. It is suggested that a research coordinator should be hired in academic health centers to optimize clinical research in the primary care setting.

The next step is to integrate the activity tracker into the electronic medical record, as suggested by our team. Shannahan et al [38] demonstrated that the patients’ activity tracker data can be embedded within visits with primary providers to personalize recommendations and that patient-physician information sharing is feasible. They conclude that activity trackers may foster patient-physician communication regarding PA, but infrastructure and resources are needed [38]. Bliudzius et al [39] stated that the data from physical monitoring systems and external medical devices should be integrated into the medical record system as these are essential in clinical work. Moreover, system integration is useful to make detailed analyses and have a global or clear picture of the patient’s health [39]. We strongly believe that patient PA data and other cardiometabolic parameters are essential in clinical work, especially in primary care settings. It helps health care practitioners to review and use patient data collected, understand how the patient feels in real-life situations, adhere to the physician’s or health professional team’s recommendations, and thus solve problems faster [39]. In order to implement the technology on a larger scale, it should be important to obtain consent and explore the patients’ stress over having their data shared with a third party. As reported by Hodgson et al [4], there is future scope for using Fitbit activity trackers to support active lifestyles in adults diagnosed with type 2 diabetes. More detailed discussion with health care professionals could identify methods of integrating activity trackers into the care of patients.

This study has a number of limitations. The first limitation is the small sample size, which could compromise the power of the study. However, the sample size was determined for the participant according to the availability of the equipment, and for the team members, it was determined according to their involvement in the pilot study. We believe that we have gathered sufficient helpful information to draw general recommendations to optimize the implementation of this technological tool in primary care. It is also important to report that some satisfaction may be tracker-dependent. The use of more sophisticated devices (eg, AppleWatch) may increase the satisfaction but also cause some other technical issues (eg, the download of the data). Therefore, some conclusions should be reformulated as “device-related,” since changing the used device may give different results. Another limitation is the relatively short-term follow-up period of 3 months, which was selected in the context of a pilot study to prevent an increased withdrawal rate in the control group and to align with our budget. However, we are aware that a longer duration may have been beneficial for incorporating lifestyle habits and also for a long-term comparison with the control group. Moreover, a longer duration could be interesting to observe the tendency of motivation over time because the highest step average was recorded after the follow-up phone call and the lowest weekly step average was recorded at week 12 at the end of the intervention. A long-term study will be interesting to observe the motivation tendency and find solutions to optimize and maintain PA motivation with an activity tracker over time. It is interesting to point out that most studies, even pilot studies, have recommended a follow-up duration of at least 6 months and preferably 12 months or longer. Similar studies had a follow-up period of 3 to 6 months [1,2,40]. A meta-analysis of activity trackers in adults with cardiometabolic conditions reports a median duration of 17 weeks and up to 18 months [37]. Another study on young adult cancer survivors found that a 12-week Fitbit and Facebook-based physical intervention was feasible for this population and had promising effects on reducing sedentary time [41].

In conclusion, patients with type 2 diabetes were satisfied with the activity tracker and felt that it incited them to stick with their PA program once the study was over. According to the research team, its implementation is feasible in primary care, but some challenges remain to having this technological tool in clinical practice. From a clinical innovation perspective, it would be interesting to find a way to synchronize relevant activity tracker data to the electronic medical record to optimize the collaboration between patients and health professionals in a primary care facility.

Based on a patient partner’s idea and his continuous involvement, this project showed that laypersons have an important role in implementation research by informing the design of realistic interventions and optimizing their feasibility. Moreover, health researchers, clinicians, and other health care professionals need to clarify the opportunities to integrate digital technologies into public health to maximize their potential to improve public health outcomes and patient care.