A descriptive cross-sectional survey with a comparison group was conducted. We included a comparison group as we wanted to compare the outcomes of veterans with those of able-bodied controls (matched to age and sex) living in the same geographical location, having similar socioeconomic and lifestyle contexts, and having access to similar health care resources.

Qualitative semistructured interviews were conducted using a phenomenological approach. This approach was chosen to encourage the identification of broader emerging themes that crosscut the diverse health, social, societal, and individual factors known to affect engagement in and effectiveness of rehabilitation with regard to PA and QoL. Interviews were designed based on methods described by Creswell [30] in planning and conducting qualitative research and published studies focused on factors influencing PA participation among individuals with chronic disabilities, including LLA [16,19,31].

We identified potential veterans with LLA (group 1) from the “Disabled Category Registry.” We aimed to include 85 participants in each of the groups so as to adequately power the comparison of each outcome [32]. We ensured representation from veterans across all 5 districts, selecting participants proportionally using a stratified random sampling procedure [33,34]. Participants for the comparison group (group 2) were identified from the same village or a neighboring village of their group 1 counterparts using the voter registration list.

We selected veterans who had LLA due to an injury on the battlefield and were living in the community. To ensure that the participants had the required functional level for the proposed CBPR intervention, we included only participants who had unilateral LLA and used a prosthetic limb for walking and standing activities. Veterans older than 70 years and those with comorbidities that interfered with their function beyond that of unilateral LLA (eg, dependence on renal replacement therapy) were excluded.

For the interviews, we purposively selected participants from group 1, ensuring participation from all 5 districts with regard to transfemoral and transtibial amputations to assess the needs and understand the perspectives of individuals with different functional levels of mobility after amputation [35,36].

The self-administered SF-36 [37] and International Physical Activity Questionnaire (IPAQ) long-form survey [38] were used to assess QoL and PA participation, respectively. The SF-36 is widely used to measure QoL in terms of physical (physical component summary [PCS]) and mental or emotional (mental component summary [MCS]) components, each expressed as a value between 0 and 100, with a high score representing a better QoL [37]. The IPAQ measures the frequency (days per week), duration (minutes), and level of intensity (vigorous, moderate, walking, or sitting) of PA during the last 7 days [38]. Both questionnaires have established psychometric properties making them ideal for use in the LLA population [25,39,40] and have been validated for use in the Sinhalese population previously [41,42].

Initial contact with the participants was made through Grama Niladhari (GN) and officers of societies of amputee veterans (eg, Ranaviru Sansadaya). GN is a Sri Lankan public official appointed by the central government to carry out administrative duties in a GN division (geographic region), which is a subunit of the divisional secretariat.

Participants of groups 1 and 2 were met by 2 research team members (AW and Dasun Isurinda). AW explained the research, provided participants with study information, and sought consent. AW is a trained physiotherapist fluent in Sinhala, with over 6 years of experience working in both clinical and research capacities within community settings in Sri Lanka. Dasun Isurinda is a practicing physiotherapist with more than 5 years of experience in both inpatient and community physiotherapy settings. The SF-36 and IPAQ were available to participants in paper form in the local language (Sinhala). AW and Dasun Isurinda remained with the participants during the survey completion to answer any questions the participants may have regarding the self-assessment.

The interview guide was developed using the Theoretical Domains Framework (TDF) [43]. This included knowledge about PA or exercises and exercise programs, intentions for participating in PA, environmental context and resources, emotions on life with LLA, and reinforcement through support. The guide was translated by a bilingual research team member (AW) and then back-translated and checked for accuracy by a second researcher (SJ). It was piloted with veterans with LLA who were not included in the final analysis. The pilot resulted in the simplification of the question format. The final survey can be found in Multimedia Appendix 1.

A total of 25 interviews were conducted in the language preferred by the participants (Sinhala) and lasted between 30 and 40 minutes. All the interviews were conducted by the author AW at the residence or home of each participant.

All the statistical analyses were performed by author DGD (a qualified statistician; independent of participant allocation and data collection) using STATA/IC for Mac v16.1 (StataCorp). The normality of data distributions was tested with the Shapiro-Wilk test, and data are summarized as mean (SD), median (range), or number (percentage), as appropriate. The Mann-Whitney U test and chi-square test were used to evaluate comparisons between groups for continuous and nominal variables, with a significance level of .05.

Data from the IPAQ were processed and reported according to the Guidelines for Data Processing and Analysis of the IPAQ [44]. In the IPAQ, PA is defined in terms of the metabolic equivalent of task (MET) minutes per week, and the questionnaire assesses PA participation in walking, moderate-intensity, and vigorous-intensity activities across 4 domains: work, transport, domestic and garden, and leisure. We computed the PA participation of groups 1 and 2 separately for each of these domains and calculated the total PA level by adding them together. Finally, the level of PA was classified as either sedentary (<600 MET-minutes/week), low (600-3000 MET-minutes/week), or sufficiently active (>3000 MET-minutes/week), based on the total MET-minutes/week [44] for both groups.

The findings of the qualitative study were reported using the Consolidated Criteria for Reporting Qualitative Studies guidelines [45]. Findings were thematically analyzed using the Consolidated Framework for Implementation Research (CFIR). The aim of using the CFIR was to identify the different organizational levels to which the identified barriers, facilitators, and expectations for a future CBPR intervention belong, in order to gain further insights into the effective design and implementation of the intervention considering each organizational level. The CFIR is a pragmatic meta-theoretical framework that helps to identify determinants of a health care intervention implementation with consideration for context, the complexity of the intervention, individual characteristics, and organizational or system-level factors that may facilitate or inhibit implementation [46-48].

Thematic analysis was used to identify emerging themes from the interview responses [30]. Responses were initially reviewed independently by 2 researchers (AW and Nilu Dullewe, both qualified health care professionals trained in qualitative methods and fluent in the Sinhala language) who read through all the verbatim transcripts to inductively code sentences and keywords. Emerging themes were then codified using the domains of the CFIR. These were then reviewed by both researchers, duplicates were removed, and emergent themes were refined. Any disagreements that developed during the analysis were discussed, and if needed, these were further reviewed by the author AB, a clinical researcher with experience of both the Sri Lankan health care setting and the methods used for analysis.

QoL scores by SF-36 domains are presented in Table 2. The median cumulative scores of physical health (PCS) and psychological well-being (MCS) were significantly lower in group 1 than in group 2 (P<.001). The difference in the PCS score had a large effect size (r=0.5), while the difference in the MCS score had a medium effect size (r=0.3). For group 1 participants, the poorest QoL scores were related to general health (median 45, IQR 55-35) (Table 2).

In the comparison of QoL outcomes between different amputation levels, only the “general health” domain (under PCS) showed a significant difference, with a lower value for veterans with transfemoral amputation (P=.009; Multimedia Appendix 2).

The total PA level was significantly lower in group 1 than in group 2 (P<.001), with a medium effect size (r=0.3). Participation in walking, moderate-intensity, and vigorous-intensity activities was lower in group 1 than in group 2, with a significant difference in walking (small effect size of r=0.2; P=.004) and vigorous-intensity PA (medium effect size of r=0.3; P<.001) (Table 3). Among 79 veterans, 59 (75%) did not meet the recommended PA level (>3000 MET-minutes/week). Moreover, the “sedentary” level was noted in 43% (34/79) of participants in group 1 and 12% (10/82) of participants in group 2 (P<.001) (Table 4).

Of the 79 participants with LLA, the majority engaged in moderate-intensity PA inside the house (49/79, 62%) and in the yard (30/79, 38%). The least participation was in cycling for transport (5/79, 6%) and vigorous PA (recreation, sport, or exercise) in leisure (6/79, 8%) (Multimedia Appendix 3).

When considering the amputation level, participation in walking was significantly lower among veterans with transfemoral amputation than among those with transtibial amputation (P=.01), and 4 out of the 5 participants (80%) with transfemoral amputation had PA levels below the recommended guidelines (Multimedia Appendix 2).

Participants lacked structured CBPR programs and community-based follow-up care from rehabilitation health providers, primarily due to the absence of formal community rehabilitation services like physiotherapy. They believed that having a CBPR program upon discharge from institutional care would have increased their PA participation.

Prosthetic limbs were the only means of ambulation for the veterans included in this study, and they are needed to engage effectively in PA. Veterans are given free prosthetic legs by nongovernmental organizations to support their independent mobility. However, this service was not available all the time and was only available to a few of the participants. People who received this service had the opportunity to replace a worn-out prosthesis with a new one.

Continuous financial support from the government in the form of a monthly salary and disability allowance relieved participants of the burden of earning money for their households, enabling them to dedicate ample time to PA and exercise. However, this support led to sole reliance on the allowance, discouraging them from pursuing any occupational opportunities. This was connected to reduced participation in work-related PA and reduced motivation for PA, which has been described under the subsection “Individual Characteristics.”

Veterans resided in designated villages allocated for army veterans, providing a peaceful environment with ample space for PA like walking and gardening. Living among peers with similar mental and physical states fostered an inclusive environment, where disabilities were not emphasized, encouraging frequent sharing of thoughts and experiences. Additionally, kinship with family and associated competing responsibilities hindered their engagement in PA.

Participants did not have adequate space and equipment to engage in exercise and PA. They believed that exercise would not be effective without proper exercise equipment. The poor functionality of the prosthetic leg combined with skin wounds resulting from its incorrect fitting posed challenges for participating in PA, particularly walking activities.

Acquiring proper knowledge and training is crucial for successful and effective engagement in PA. However, participants expressed a lack of access to professionals or services to seek information and guidance on performing exercises at home.

Participants expressed uncertainty about what to expect in terms of recovery upon discharge from inpatient care. They lacked an understanding of the importance of ongoing exercise engagement for their recovery, with some believing that exercises would not contribute further to their progress. Instead, they perceived activities, such as household chores, gardening, and walking to nearby shops or houses, as sufficient for maintaining a healthy life.

Owing to their active lifestyle before the injury (heavy physical training in the army and representing army sports teams), participants believed that they were familiar with the basic exercise principles. This helped them to engage in at least a few exercises at home even without proper guidance or follow-up.

Veterans reported various health comorbidities, including back pain, knee pain, diabetes mellitus, and hypertension, which affected their ability to engage in PA. Veterans with transfemoral amputation perceived lower PA abilities compared to those with transtibial amputation, and they anticipated a worsening situation with age. In contrast, some veterans associated their current physical state positively with engaging in PA. One reason they mentioned was being an active prosthetic user, which made them independent in walking. As they joined the military service at 18-24 years of age and got injured at a young age, their relative age at the time of injury was seen as a facilitator to recovery.

Participants considered engaging in PA and exercise as an extra burden, requiring them to modify their usual lifestyle. Some expressed a lack of motivation for any form of PA, including walking for daily tasks. In contrast, for some participants, consistent engagement in PA was considered crucial among individuals with disabilities. It was seen as a lifelong requirement rather than a lifestyle choice to improve functional levels and reduce the risk of health issues, such as diabetes mellitus and heart disease. They expressed motivation to increase their activity levels and independence, aiming to avoid dependence on family members, including that related to the incidence of chronic health conditions.

Participants expected the CBPR program to be tailored to their disability, with exercises matching their current functional levels. Exercises to prevent deterioration in existing health, notably back pain and knee pain, were a particular priority for participants with chronic comorbidities.

Veterans held a favorable perception toward CBPR, perceiving it as easily adaptable to their needs. They expressed a preference for engaging in rehabilitation programs either at home or within their local community, as opposed to attending outpatient clinics at hospitals. This preference was associated with perceived benefits, such as reduced travel burden, lower associated costs, and minimized disruptions to their daily lives.

Veterans generally suggested a program with 2 to 3 sessions per week, lasting 20-30 minutes each. They favored simple functionally oriented exercises that could be easily incorporated into daily activities, with a preference for specific exercises, such as leg muscle strengthening.

Veterans preferred using exercise equipment only when necessary, considering constraints like the lack of equipment at home and financial limitations. They showed interest in using community spaces, such as playgrounds and meeting halls, for group rehabilitation sessions when home space was insufficient. Group participation was favored for the opportunity to learn from one another during the program.

Participants stressed the importance of an awareness program led by experts in the field to precede the implementation of a future CBPR program, with the aim of ensuring maximum engagement of veterans in the CBPR program.

Veterans highlighted the necessity of key contact from both veterans and rehabilitation providers for each village. This is to communicate the necessary information smoothly and get advice when necessary.

The veterans were members of formal societies like “Ranaviru Sansadaya,” which are associated with enabling participants to stay united as one group of army veterans and connecting them with external organizations to receive help. They anticipated that delivering the program through these societies would help initiate and continue the program successfully.