This quality improvement study aimed to inform the development of an activity monitoring program to complement a clinical intervention for veterans with GWI. The VA New Jersey Healthcare System Institutional Review Board determined that this quality improvement project met criteria for quality improvement and was considered non-research; therefore, formal institutional review board review was not required. The CONSORT (Consolidated Standards of Reporting Trials) extension for pilot and feasibility studies was used for reporting (Multimedia Appendix 1) [17,18].

Veterans with GWI who completed a comprehensive evaluation at the WRIISC-NJ were invited to participate in a 26-week virtual health coaching intervention program. Two cohorts were enrolled: April 2023 to September 2023 and October 2023 to March 2024. GWI diagnosis was determined clinically through WRIISC-NJ specialty evaluation.

Participants received a smartwatch (Fitbit Sense 2) as part of a VA device distribution initiative [19] as well as a user guide. A videoconferencing orientation provided an overview of device setup, smartwatch capabilities (eg, tracks physical activity and heart rate), program expectations, risk mitigation (ie, taking breaks from wearing the smartwatch to mitigate skin irritation), and data privacy and sharing through a third-party Health Insurance Portability and Accountability Act (HIPAA)-compliant platform (Fitabase). Verbal consent for data sharing was obtained. Participants did not receive monetary compensation but could keep the smartwatch after program completion. Feasibility and acceptability were assessed using smartwatch wear-time data, monthly electronic surveys, and a semistructured interview.

Participants were instructed to wear the smartwatch, keep it charged, sync data using the manufacturer’s mobile app at least once every 7 days, and agree to share data with the project team. These procedures were intended to reflect typical real-world use rather than impose artificial workflow constraints on participants. No additional instructions were provided regarding when or how long to wear the device beyond wearing it as tolerated. The program was designed to capture participant behavior in a free-living environment. A nurse educator reviewed battery status and “last synced” data twice per week and contacted participants if more than 6 days had elapsed without data syncing. During the postprogram session with the clinical director, participants were presented with a report summarizing their average weekly step count and sleep duration.

Program feasibility was assessed using the number of participants recruited and enrolled, average daily wear-time percentage, and duration of smartwatch use (weeks) prior to discontinuation. Adverse events related to smartwatch use were recorded. Program acceptability was assessed using two items from the monthly Qualtrics survey: (1) the extent to which the smartwatch has helped them achieve their wellness goals (0=not at all; 10=definitely) and (2) the likelihood of recommending the program to another veteran by using a modified 1-item net promoter score (0=not at all likely; 10=extremely likely) [20].

Questions focused on smartwatch use, perceived benefit, and anticipated continued use. These questions were loosely adapted from interview guides from other interventions conducted by the senior researchers (GA, SM) that had included questions about acceptability of technology use. Questions included, “How do you use your Fitbit?” and “Do you think you’ll keep wearing the watch after the program has ended? Why or why not?” A nurse educator (author PM) trained in motivational interviewing contacted the participants by phone after the program midpoint to conduct a semistructured interview regarding smartwatch use. Two attempts were made to reach each participant. Interviews were scheduled to coincide with other program evaluation activities when possible. Interviews lasted approximately 10 minutes, and responses were recorded using a Microsoft Teams form.

Twelve individuals enrolled in each of the two cohorts (n=24). Three later withdrew from the health coaching program, and 1 declined to participate in the activity monitoring program, resulting in a final sample of 20 participants (mean age 49 years; 7/20, 35% female; 19/20, 94% non-Hispanic White; 11/20, 55% first-time smartwatch users; Table 2). Nineteen participants completed at least one of the 6 monthly surveys, including 11 who completed all 6 surveys, 4 who completed 5 surveys, and 4 who completed 2-4 surveys. All participants completed a semistructured interview.

Eighteen participants responded at least once to the question, “To what extent did the Fitbit program help you achieve your wellness goals?” Final responses ranged from 0 (not helpful at all; n=1), 1-5 (n=5), 6-8 (n=9), to 9-10 (very helpful; n=3). Among the 17 participants who completed at least two surveys, the largest observed change was from 3 to 10. All but one participant showed an increase from baseline to final response.

Nineteen participants responded at least once to the recommendation question. Final ratings ranged from 0 to 6 (n=4), 7 to 8 (n=8), and 9 to 10 (n=7). The final net promoter score was 15.8, indicating a favorable score. All but 4 participants reported increased recommendation scores over time.

Twelve participants who rated the device as helpful (rating greater than 5) also indicated they would recommend the smartwatch to other veterans (rating greater than or equal to 7). Four of 6 participants who did not rate the device helpful would still recommend the device to other veterans.

Total wear time ranged from 8 weeks to 26 weeks (median 26, IQR 18-26). Twelve participants (60%) wore the smartwatch for the full 26 weeks. The remaining 8 participants discontinued use between 8 and 12 weeks (n=3), 16 and 18 weeks (n=4), and 23 weeks (n=1).

Average daily wear-time completeness ranged from 10% (2.4/24 hours) to 96% (23.0/24 hours) across participants. Among those who completed 26 weeks, daily completeness ranged from 76% to 96% (18.2-23.0/24 hours; mean 87% [20.9 hours], SD 6% [1.0 hours]). Prior to device discontinuation, 6 participants demonstrated similarly high completeness (mean 85% [20.0 hours], SD 8% [2.0 hours]).

Of the 11 first-time smartwatch users, they were more likely to be younger than 50 years (8/11, 73%) and female (5/11, 45%). First-time users showed higher rates of early discontinuation (6/11, 55%) and low daily completeness (2/11, 18%). Figure 1 shows the weekly distributions of daily wear-time completeness across participants. In most weeks, wear-time distributions were negatively skewed, with a subset of participants demonstrating substantially lower wear time than the group median. By approximately week 18, this pattern diminished as participants who discontinued use had stopped contributing data. The remaining 12 participants who wore the smartwatch for the full 26 weeks generally maintained consistently high weekly wear time.

There was one adverse event reported. One participant reported minor skin irritation related to the watch band, which resolved after switching to a hypoallergenic watch band.

Findings from the semistructured interviews are organized below by domain from TFA. Representative quotes are provided in Table 3.

In this program, smartwatch-based activity monitoring was incorporated into a virtual health coaching program for veterans with GWI with sustained use and participation. High rates of sustained participation and daily wear-time completeness indicate that most participants were willing and able to engage with activity monitoring over an extended period. Favorable survey scores and qualitative feedback further support the acceptability of this approach. From a human factors standpoint, the combination of sustained wear, low reported burden, and high self-efficacy suggests that the device-user interaction was well aligned with participants’ capabilities and daily workflows. Program observations support the feasibility and acceptability of incorporating activity monitoring into health coaching programs for veterans with GWI.

This paper presents findings from the first known activity monitoring program specifically for veterans with GWI. Prior studies have shown that wearable activity trackers can be implemented in populations with chronic disease and are commonly used to support physical activity and self-monitoring, which is consistent with the patterns observed in this program [24-28]. Using a smartwatch for activity monitoring may help inform the design of personalized symptom-management strategies and may provide signals regarding treatment response. This may be applicable for veterans with GWI with decreased cognitive function, since the smartwatch may serve as a cognitive support tool to reduce tracking burden, create accountability, increase awareness of how behavior can affect health, and prompt discussions with care team [27].

Assessing feasibility and acceptability are conventional practices for researchers when introducing new program components, particularly in unique clinical populations such as veterans with GWI who have complex and variable health needs. A feasibility pilot would also explore barriers and pain points related to technology use prior to broader implementation. Technology acceptance frameworks such as the Technology Acceptance Model describe how perceived usefulness and perceived ease of use shape attitudes toward technology and influence actual use [29]. Although our qualitative analysis was guided by TFA, these constructs align conceptually with the Technology Acceptance Model and helped clarify why participants did or did not accept the activity monitoring program.

Prior studies have identified factors such as social features and engagement strategies that influence the acceptability of mobile health components, including social connectivity [30], gamification [31], and ability to choose among available devices [32]. Device-agnostic platforms may increase flexibility and user choice; however, they can also introduce challenges related to data integration and standardization across devices [32,33]. Issuing a single device to all participants can simplify data management and support consistent clinical feedback, but it may also introduce tension for participants who prefer another brand of smartwatch. Our findings reflected this complexity. Several participants expressed loyalty to other smartwatch brands; however, most continued to wear the program-issued device, suggesting that brand preference alone did not determine acceptability of the activity monitoring program. Similar to prior digital health programs, sustained engagement tends to decline over time in app-based and wearable-based interventions [34,35].

Acceptability in this program appeared to be multifaceted and strongly shaped by clinical context. Survey results suggested that perceived helpfulness was closely aligned with willingness to recommend the program, supporting overall acceptability. At the same time, qualitative data indicated that some participants viewed the smartwatch as less useful, preferred a different brand, or questioned specific device features. Despite these reservations, many continued to wear the program-issued device because it was embedded within the larger health coaching program. For these participants, participation in the core health coaching intervention and their commitment to the program appeared to outweigh dissatisfaction with the specific device. This suggests that acceptability of the activity monitoring component was influenced not only by the smartwatch itself but also by veterans’ motivation to fully engage in the broader health coaching program.

This quality improvement project was designed to assess feasibility and acceptability rather than effectiveness; therefore, causal inferences cannot be made. Smartwatch data were collected in an uncontrolled free-living environment and could not be independently validated. Although the sample size was modest, it was sufficient to detect common feasibility challenges in pilot work [36]. Participants were recruited from a single VA center, which may limit generalizability to the broader population of veterans with GWI. Because activity monitoring was implemented as part of a health coaching program, our findings may not generalize to a standalone activity monitoring program or to other populations.

The growing availability of smartwatches creates an opportunity to support symptom self-management for veterans with GWI. In this program, activity monitoring was implemented to complement a 26-week virtual health coaching program. The program demonstrated high feasibility and acceptability within this clinical setting, supporting activity monitoring as a feasible way to augment patient education and ongoing support for veterans with GWI. Program acceptability appeared to be influenced by both the usability of the device and its integration within the existing health coaching program.

When conducting feasibility work, programs should be designed to reflect the specific setting and population of interest. Collecting both quantitative and qualitative data provides important context for understanding use and engagement. Assessing feasibility and acceptability can then be used to inform implementation strategies tailored to supporting sustained participation in the program.