In Canada, over 65,000 people were estimated to be living with HIV at the end of 2022 [1]. More individuals are living longer and aging with HIV; globally, persons living with HIV aged 50 years or older increased from 5.4 million in 2015 to 8.1 million in 2020 [2]. As people living with HIV age, they can experience health-related consequences of HIV, including an increased risk of cardiovascular disease, diabetes, bone and joint disorders, neurocognitive disorders, and certain cancers [3-5]. Rehabilitation, such as physical therapy and occupational therapy, can mitigate disability associated with HIV and multimorbidity, such as fatigue, pain, cognitive impairments, body composition changes, and employment challenges [6-9]. Exercise is a rehabilitation intervention and self-management strategy that can address disability and improve or sustain the health of individuals with HIV. Systematic review evidence indicates that exercise is safe and can lead to benefits in cardiopulmonary fitness, strength, weight and body composition, and psychological outcomes for people with HIV [10,11]. However, this evidence primarily involved supervised interventions by physical therapists and exercise physiologists that can be costly and unsustainable [10,11]. With up to an estimated 73% of people living with HIV classified as sedentary or physically inactive, it is essential to consider exercise interventions that are accessible and practical for people with HIV to sustain over the long term [12].

Community-based exercise (CBE) is a promising approach that may help people living with HIV increase their engagement in exercise and physical activity [13-16]. Exercise is purposeful, planned, and repetitive bodily movement that has the objective of improving or maintaining physical fitness, while physical activity is any bodily movement that results in energy expenditure [17]. Persons living with HIV perceive benefits from exercise and physical activity as a component of health promotion, with the potential to improve physical function and mental health and reduce isolation [18]. CBE can be provided under the guidance of a health or fitness instructor, with the aim of encouraging consistent exercise and physical activity within the community [13,14,16]. CBE enhances the benefits of traditional exercise as it can improve social interaction, provide support and motivation, and promote emotional, cognitive, and behavioral self-management strategies to help individuals manage health challenges [14,19,20]. CBE has been associated with engagement in physical activity and improvements in health outcomes for adults living with HIV [19,21]; however, the factors influencing the implementation and uptake of CBE are less clear.

Using the Consolidated Framework for Implementation Research (CFIR), Cooper et al [22] conducted a qualitative systematic review of CBE interventions to explore barriers and facilitators of implementing CBE interventions in the general population [22]. Barriers included lack of participant motivation, insufficient resources allocated for implementation, the complexity of the CBE intervention, and lack of participant knowledge about the intervention [22]. Facilitators included the flexibility to adapt the intervention to the context, intervention compatibility with participants, and participants having a positive perception of those implementing the intervention [22]. The authors concluded that making minor changes to tailor an intervention while maintaining core components, such as clear implementation strategies and coordination of delivery, can facilitate the implementation of a CBE intervention.

Members of this authorship team assessed the impact of an in-person CBE intervention with adults living with HIV at the YMCA in Toronto, Ontario, Canada [21]. While the team reported improvements in cardiovascular health, strength, flexibility, and self-reported physical activity, barriers to taking part in CBE were documented among women, persons who are gender diverse, and persons who did not live geographically close to the YMCA [21]. Persons with HIV may experience obstacles to exercising in traditional gym environments, including interpersonal, financial, and geographical barriers, self-image issues and stigma, and difficulty initiating exercise following periods of inactivity [23,24].

Telerehabilitation, and more specifically telecoaching, is a potential strategy for implementing CBE with persons living with HIV. Telerehabilitation is the delivery of rehabilitation programs or services via technology, such as phone and video calls, web-based platforms, or mobile apps, used initially to increase efficiency, quality of care, and access to health care for populations with geographical [25,26], financial, or time-limiting barriers for persons living with HIV [27]. Telecoaching is a component of telerehabilitation that uses technology for remote supervision, guidance, and communication of an exercise program to improve physical activity levels and manage symptoms for people living with chronic diseases [28-33].

We established an online telecoaching CBE intervention with the aim of enhancing engagement in physical activity among adults living with HIV in Toronto, Canada [34]. The intervention was developed based on foundational considerations for developing and implementing online CBE with adults living with HIV that included: (1) person-specific considerations (episodic nature of HIV, stigma, and HIV disclosure), (2) accessibility of the program, (3) program delivery and technology, (4) attributes of program personnel, (5) program content and design, and (6) building community and peer support [35].

Fidelity of implementation (FOI) is important to consider in online CBE interventions to develop practical interventions that lead to sustained physical activity. FOI, or integrity, is the degree to which programs are implemented as intended by the program developers [36-38]. The National Institutes of Health Behavior Change Consortium (BCC) treatment fidelity framework specifically refers to treatment fidelity and the link between training the providers, delivery, receipt, and enactment of the intervention for research involving health behavior interventions [39]. The fidelity with which an intervention is implemented can influence its level of success and uptake, and the impact on the outcome, with high fidelity promoting improved outcomes. However, it is important to consider planned adaptations to the intervention to optimize treatment fidelity while acknowledging the need to accommodate practical needs, study demands, and provider and participant burden, requiring pragmatic approaches in real-world settings [39]. Lack of fidelity can lead to failure in intervention implementation in a real-life context [40]. Hence, reporting fidelity is essential to determine the credibility, validity, and replicability of intervention study findings [41]. Despite the importance of understanding intervention fidelity, lack of knowledge or understanding about assessing fidelity may be barriers to assessment and reporting [42]. While quantitative and qualitative methods are recommended to comprehensively assess fidelity, this guidance is not consistently followed [39,42-44]. To our knowledge, there is limited evidence assessing the FOI of exercise interventions for persons living with HIV. Assessing the FOI of exercise interventions in the context of HIV is important because (1) sustained exercise can mitigate HIV-related disability, and (2) people living with HIV can face unique barriers and facilitators to exercise, such as episodic disability, which can influence the FOI of exercise interventions [45].

Our aim was to describe the FOI of a 6-month online telecoaching CBE intervention with adults living with HIV from the perspectives of participants (adults living with HIV) and personal trainers.

Findings from this study may help inform policy and programming for broader scale-up and optimal CBE implementation that can be adopted by fitness centers and community-based organizations to proactively prevent multimorbidity and improve health outcomes among persons living with HIV.

We conducted an observational longitudinal study involving (1) online structured interviews with adults living with HIV who participated in a 6-month online CBE intervention and (2) coaching logs of training sessions completed by personal trainers. Details of the study protocol have been previously published [34]. Figure 1 provides an overview of the CBE intervention and fidelity outcomes.

We included adults living with HIV who participated in the telecoaching study and personal trainers from the YMCA who were involved in supervising personalized online exercise sessions with the study participants [34]. Eligibility criteria for the telecoaching study were adults (aged 18 years or older) who considered themselves safe to engage in exercise as determined by the self-administered Physical Activity Readiness Questionnaire (PAR-Q) and had access to an electronic device with an internet connection and a webcam, as well as space at home for exercise [34]. We recruited potential participants living with HIV through community-based organizations, an HIV clinic in Toronto via the Ontario HIV Treatment Network Cohort Study (OCS), and participants from a prior CBE study who agreed to be contacted about future research [21,46].

Prior to the CBE implementation, fitness instructors (certified personal trainers and staff at the YMCA) took part in workshops on HIV, rehabilitation and exercise for adults living with HIV, goal setting, and research procedures [47]. We piloted the setup of the technology, including (but not limited to) Zoom software [48], a webcam, YMCA membership, web-based questionnaire software, and the Fitbit app [49] with the participants living with HIV and fitness instructors.

The telecoaching online CBE intervention included a 6-month structured intervention phase followed by an independent 6-month exercise phase [34]. During the 6-month intervention phase, participants were asked to engage in the following three components of the intervention: (1) an exercise regimen of aerobic, resistance, balance, and flexibility exercise occurring thrice weekly at moderate to vigorous intensity for ≥60 minutes per session; (2) attend online biweekly individually tailored one-on-one exercise sessions with a certified personal trainer from the YMCA, involving supervision and progression of aerobic, resistance, balance, and flexibility exercise; and (3) engage in online monthly group educational sessions on sleep, nutrition, goal setting, cognitive health, chronic pain, and other topics related to self-management, health, and physical activity while living with HIV, delivered by members of the research team. Each participant met with their personal trainer prior to the intervention to tailor the intervention to their goals, abilities, and interests [50]. Participants received a YMCA membership allowing them to access online YMCA classes through the YMCA SweatforGood app. Personal trainers uploaded the individualized exercise plans to the SweatforGood app, allowing participants to access, review, and track their exercises [34].

After the 6-month intervention, participants entered a 6-month follow-up independent exercise phase, during which they were asked to continue with thrice-weekly exercise. They continued to have access to the YMCA online membership, providing access to group online classes, and the SweatforGood app, but the study no longer provided one-on-one personal training sessions or monthly educational sessions.

Participants were given home exercise equipment (Therabands and a wooden step) and a wireless physical activity monitor (Fitbit Inspire 2) to monitor and encourage exercise during both the 6-month CBE intervention and the 6-month follow-up period. Participants documented their physical activity through a weekly online exercise log. Participants were given the Fitbit Inspire 2 and exercise equipment to keep at the end of the study as a token of appreciation for their participation. For more details, see the published protocol [34].

We aimed to recruit 40 adults living with HIV and 30 to complete the intervention, enabling us to achieve our primary study aim of assessing the impact of the CBE intervention [23,51].

We defined overall FOI of the CBE intervention as achieved if the following 4 criteria were met by ≥80% of the participants in the study: criterion 1, attended ≥80% (11/13) of their 13 training sessions; criterion 2, engaged in (at least) thrice-weekly exercise in the week of the personal training session ≥80% of the time over the 6-month intervention, as recorded by the personal trainers on the coaching logs; criterion 3, reported “complete criteria met” for ≥80% of the fidelity elements (at least 16 of 20 items) at months 2 and 6, as reported in the interviews with participants living with HIV; and criterion 4, engaged in a combination of aerobic, strength, balance, and flexibility exercises during the supervised coaching sessions ≥80% of the time over the 6-month intervention, as reported by the personal trainers in the coaching logs. Our criteria are based on the literature, whereby 80% to 100% is considered high fidelity, and ≤50% defines low FOI [54-56].

This study was approved by the University of Toronto Health Sciences Research Ethics Board (#40410). Participants provided informed verbal consent to participate in the study, which was documented on the consent form. Participant data were anonymized and stored on a secure server at the University of Toronto. Participants did not receive compensation for their participation in the exercise intervention. They were able to retain their exercise equipment and wireless physical activity monitor at study completion.

Adults living with HIV who initiated the intervention and had at least 1 completed coaching log were included in the study. Of the 32 participants, 18 (56%) completed the 6-month intervention and had an FOI interview assessment. A total of 29/32 (91%) participants who completed at least 1 fidelity interview were included in the FOI analysis. Further, 16 (50%) participants completed the FOI assessment (interviews) at both months 2 and 6; 11 (34%) completed the assessment at month 2 only, and 2 (6%) completed the assessment at month 6 only. Of a potential 64 fidelity interviews, 45 (70%) were completed. At month 2, a total of 84% (27/32) of assessments were completed, and at month 6, a total of 56% (18/32) of assessments were completed.

Of the 32 participants at baseline, the majority were male and living with 2 or more concurrent health conditions in addition to HIV. Table 2 provides characteristics of participants enrolled in the study at baseline, month 2, and month 6.

Of the 32 participants, a total of 8 (25%) who initiated the intervention completed all 13 coaching sessions, 7 (24%) of whom were among the 29 participants who completed an FOI assessment interview. We received responses to 72% (298/416) of the coaching logs out of the total possible for 32 participants living with HIV to determine FOI from the perspectives of personal trainers.

Among participants who completed the intervention and FOI assessment at month 6 (n=18), FOI was achieved for 1 of the 4 criteria, whereby 83% (95% CI 66%-100%) of participants who completed the study engaged in ≥80% of biweekly online personal training sessions. However, fidelity criteria were not met for independent exercise thrice weekly (50%; 95% CI 27%-73%), participant-reported fidelity criteria at month 6 (“complete criteria met” for at least 16/20, 44% of items; 95% CI 21%-67%), or engagement in all 4 exercise types in ≥80% of sessions (4/18, 22%; 95% CI 3%-41%).

Within the training session, the majority of participants engaged in strength and flexibility exercises, followed by aerobic and then balance exercise. This may be attributed to strength training requiring a higher level of supervision as opposed to aerobic activity that was done independently outside the formal personal training session. In the absence of equipment for aerobic activity, such as a treadmill or stationary bike, participants may have engaged in this type of activity independently outdoors.

Fidelity was achieved for some components of the intervention. At least 80% of participants completely met 14 out of 20 FOI elements at month 2 and 6 out of 20 FOI elements at month 6, as reported in the structured interviews. The FOI items least commonly met were independent engagement in exercise (for a total of thrice weekly or more of 60 minutes or longer), attendance at weekly group-based online classes at the YMCA (8/27, 29% of participants met this criterion), and attendance at monthly online group-based educational sessions (12/27, 45% of participants met this criterion).

Low engagement in the online group-based YMCA classes and independent exercise outside of supervised one-on-one personal training sessions may be influenced by the unsupervised components of the intervention, as participants had higher engagement rates when supervised (ie, personal training sessions). This is consistent with the literature, which shows increased engagement in supervised exercise programs, with adults aged 65 years or older favoring supervised exercise programs [57]. As found within the open-ended FOI responses in this study, receiving feedback, education, and modified exercise tailored to participant preferences and abilities may be important factors driving adherence to supervised exercise [57-60]. The identification of the least-implemented intervention components can help inform future intervention design and may also inform the supports necessary for participants to sustain regular exercise routines postintervention. This may help to identify components of an intervention that require supervision and suggests the importance for future interventions to tailor CBE based on personalized preferences to promote engagement. Further, there is a need to explore population-specific factors that impact adherence to unsupervised exercise to better develop strategies that improve long-term behavioral change for adults living with HIV.

FOI criteria were not met for the group-based elements of the intervention. At month 2, a total of 45% (12/27; 95% CI 26%-63%) of participants reported attending the monthly online self-management sessions (≥80% of the time), and 22% (4/18; 95% CI 3%-41%) at month 6 (item 18 in the FOI assessment). At month 2, a total of 29% (8/27; 95% CI 12%-47%) of participants reported attending weekly online group-based exercise classes at the YMCA, and 33% (6/18; 95% CI 12%-55%) at month 6 (item 17 in the FOI assessment; Table 4). With HIV stigma being a barrier to seeking treatment, care, and health- and wellness-related programs, the lower engagement in group-based interventions may have been attributed to fear of HIV disclosure and discrimination [61-64].

Personal and environmental factors may further explain the FOI findings in this study. HIV is characterized by episodic disability, where there are unpredictable periods of wellness and illness [65]. Barriers to engaging in the intervention, as reported by participants, included illness, pain, or injury. The fluctuating nature of chronic conditions like HIV may make it challenging for participants to follow highly rigid schedules and training plans if their health and ability to engage in exercise vary day-to-day [23,51]. Perhaps our FOI criteria were too rigid and did not align with or consider the episodic and fluctuating nature of HIV. We recommend providers adapt the frequency and type of exercise to the nature of the illness to maintain the well-being of an individual living with HIV. Our findings indicate the need for flexibility with CBE interventions, enabling persons to cancel or reschedule sessions due to illness more frequently and engage in exercise at home [23,51]. We recommend that exercise interventions in this population be adaptable and considerate of the dynamic nature of the illness experience to ensure that intervention studies are pragmatic and translational to the real world [23,51,60].

While participants were provided with a YMCA online membership at no cost and access to a personal trainer biweekly, their financial situation still may have influenced engagement in the intervention. Income is positively correlated with exercise adherence [66,67]. Qualitative insights from this telecoaching CBE study reported that participants in this study highlighted cost as a limiting factor to engagement in physical activity [60]. A lower socioeconomic status may be correlated with poor exercise adherence due to a lack of leisure time. Individuals of higher socioeconomic status, on average, allocate more leisure time to exercise than those of lower socioeconomic status [66,67]. An association exists between low socioeconomic status and exercise in persons living with HIV, with barriers to exercise including caregiver responsibilities, time dedicated to childcare, unsafe neighborhoods, strict work schedules, and living in temporary dwellings [68,69]. Traditional exercise interventions may not be suitable for those living with HIV who are socioeconomically disadvantaged because of barriers related to employment, financial constraints, stigmatization, and inaccessible or lack of transportation, requiring interventions to be tailored to the needs of this population [68,69]. Collectively, it is important for health and fitness trainers to consider the environmental and personal factors that can influence exercise engagement and intervention adherence to tailor interventions accordingly. This may include tailoring modes of exercise delivery (eg, online and in-person; individual training or group-based classes) based on individual preferences.

Strengths of our approach included our multimethod data collection at multiple time points of the CBE intervention and engaging perspectives of both participants living with HIV and personal trainers involved in implementation of the intervention. FOI in our study was assessed more positively from the perspectives of the participants living with HIV (through structured interviews), as opposed to the personal trainers’ logs of the coached sessions. Our fidelity assessment from the perspectives of participants living with HIV pertained to their most recent personal training session, whereas fidelity assessment from the perspective of the personal trainers captured engagement in all coached training sessions and the independent exercise done in the preceding week. Using multiple methods to assess intervention fidelity provided a greater understanding of fidelity [70,71] and allowed for different populations to have different perspectives on a shared phenomenon [37,44,70]. Toomey et al [71] recognized that disagreements between results of different fidelity assessment approaches from providers and persons with lived experiences may occur. Some of the differences in FOI findings in our study may be due to participants’ overestimation of their performance when completing self-reported measures (2 time points), as compared to the personal training coaching logs completed by providers more frequently across the biweekly supervised sessions (13 time points). Nevertheless, our assessment captured perspectives from both persons living with HIV and the trainers at different stages of the intervention.

Our study protocol included methods for capturing the implementation of the intervention components but did not explicitly state the definition of implementation fidelity. Prior to the analysis, we defined the FOI threshold most commonly reported in the literature as achievement of ≥80% of criteria [54-56]. We considered all 4 criteria for FOI to be weighted equally; however, in reality, specific aspects of the CBE intervention may be more important (eg, attendance at biweekly training sessions) to achieve FOI than others (ie, engaging in all 4 types of exercise in the training session). Given that we did not collect fidelity data in the same manner for both target populations (persons living with HIV and personal trainers), we were unable to directly compare the influence of perspectives on fidelity assessment. Furthermore, participants were asked about fidelity based on their most recent coaching session, meaning that a number of sessions were not captured in this analysis, nor do we have the reasons for noncompletion of the FOI interviews or coaching logs. We did not report physical activity as measured by the wireless physical activity monitor, as this was beyond the scope of this study. However, as the Fitbit was a component of the intervention, our FOI assessment included Fitbit usage (item 19 on the FOI assessment), whereby 89% to 100% of participants at months 2 and 6 wore the wireless physical activity monitor regularly. The uptake and usage of the Fitbit aligns with experiences of technology uptake and usage in the telecoaching CBE study [72].

Our study is limited to participants who remained in the study and completed the FOI assessment interview at month 6. Of 32 participants, only 18 (56%) completed the intervention and an FOI assessment at month 6, resulting in an overestimation of FOI. In our sensitivity analyses, assuming participants who withdrew or were lost to follow-up did not meet the criteria, only 18 (56%; 95% CI 39%-73%) achieved the criteria of engaging in ≥80% of biweekly online personal training sessions (Table 3 and Multimedia Appendix 3), resulting in none of the 4 FOI criteria being met for the study.

Another limitation was that the majority of participants living with HIV in this study identified as males and men; hence, findings may not be transferable to individuals who identify as females or women living with HIV. Also, we did not assess the association of contextual factors such as level of income, stigma, and discrimination. Future research should consider the intersecting components of sex, gender, culture, socioeconomic status, and stigma and their influence on engagement in exercise [73].

There is a need for balancing FOI with personalized and tailored implementation of interventions for adults living with HIV [74-76]. While fidelity is important, solely prioritizing the implementation of the intervention as intended at the cost of tailoring the intervention to the population may not be the most beneficial for the population of adults living with HIV. To ensure the integrity and quality of intervention fidelity assessments while adapting the intervention to the population and its needs, it is important to define, a priori, the intervention components that may be modified to meet participant needs and circumstances. These include optimal frequency of supervised sessions (weekly vs biweekly), mode of exercise, and exercise setting to support independent exercise.

FOI was achieved for the supervised personal training components of the CBE intervention but was not achieved for independent exercise among adults living with HIV who engaged in this 6-month online CBE intervention. Future research should consider tailoring online CBE interventions to personal preferences to foster uptake and sustained independent engagement in exercise over the long term.