Chronic liver disease (CLD) is characterized by progressive deterioration of liver function persisting for more than 6 months and may result from alcohol consumption, viral infections, autoimmune disorders, or metabolic diseases [1]. According to an analysis of Global Burden of Disease 2019 data, digestive diseases continue to impose a substantial global health burden [2]. In South Korea, liver disease remains a leading cause of mortality, particularly among middle-aged men [3-6].

Metabolic risk factors, including obesity and type 2 diabetes mellitus (DM), are key contributors to the progression of metabolic dysfunction-associated steatotic liver disease [7]. These factors are associated with both liver disease progression and increased cardiovascular risk [8,9]. In contrast, lifestyle modifications—such as weight control, dietary management, and regular physical activity—have been shown to reduce the risk of liver fibrosis and slow disease progression [10-12]. These findings underscore the importance of self-management among patients with CLD and concomitant metabolic risk factors.

Self-management refers to the cognitive, behavioral, and emotional capacity required to manage the physical, psychological, and social consequences of chronic illness in daily life [13]. However, men with liver disease often demonstrate poorer self-management and a higher risk of fibrosis progression than women [14,15]. Occupational demands and constraints related to social roles may further impede self-management among middle-aged men [14]. Therefore, sustained self-management support that reflects individuals’ everyday life contexts is essential for effective liver disease management.

Family functioning and social support have been consistently associated with improved self-management adherence and better health outcomes in patients with CLD [16-18]. Greater family support is associated with higher self-efficacy, better self-management, and improved quality of life [19,20]. These findings suggest that self-management may be strengthened more effectively through family participation than through individual efforts alone.

Mobile health (mHealth) technologies have increasingly been used to support treatment adherence and self-management among patients with chronic diseases. A systematic review has reported improved adherence outcomes and generally high usability and feasibility while also noting potential barriers related to digital access and literacy [21,22]. App-based interventions have been shown to improve disease management and reduce hospital readmissions [23-25]. However, most existing interventions remain primarily patient-centered, with limited integration of family involvement [26,27], despite evidence that family participation can facilitate self-management by providing ongoing support and reinforcing behavioral changes in chronic illness care [28]. Previous studies have also reported several patient-perceived barriers to mHealth adoption, including usability issues, privacy concerns, technological limitations, and gaps in digital literacy [29]. A family-supported mHealth approach may therefore offer a promising strategy for strengthening self-management among individuals with CLD. However, the feasibility and acceptability of this approach have not been established among middle-aged men with CLD and metabolic risk factors or among their family members. A pilot evaluation is therefore needed to refine the intervention and trial procedures before conducting a definitive randomized controlled trial (RCT).

This pilot study aimed (1) to develop a 12-week mHealth coaching intervention designed to incorporate family support for middle-aged men with CLD and metabolic risk factors and (2) to evaluate the feasibility of the intervention and associated trial procedures in terms of recruitment, retention, intervention adherence, and data completeness. Preliminary app usability and changes in patient- and family-related outcomes will also be explored to inform a future definitive RCT.

This study was designed as a 2-arm, parallel-group pilot RCT with 1:1 allocation at the patient-family dyad level. The primary focus was the feasibility of the intervention and trial procedures, with patient- and family-related outcomes assessed exploratorily. This protocol follows the SPIRIT (Standard Protocol Items: Recommendations for Interventional Trials) 2025 guidelines [30]; the completed checklist is provided in Multimedia Appendix 1.

Middle-aged men with CLD and metabolic risk factors who were attending the Liver Center of Yonsei University Hospital in Seoul, Republic of Korea, and 1 family member per patient were recruited. During routine visits, attending physicians introduced the study to potentially eligible patients and referred those who expressed interest to trained research staff. With the patients’ agreement, the research staff reviewed relevant information in the electronic medical record (EMR) to confirm basic eligibility. Eligible patients and their family members who voluntarily agreed to participate provided written informed consent before enrollment as a dyad. Textbox 1 presents the inclusion and exclusion criteria for both groups.

Patient-family dyads were randomly assigned in a 1:1 ratio to the intervention or control group using block randomization with a block size of 4. The allocation sequence was generated in advance using a web-based randomization system [34] by a designated researcher who was not involved in participant recruitment or enrollment. The prespecified allocation sequence was applied to each dyad in the order in which written informed consent was obtained. Group allocation was concealed from the research staff responsible for participant recruitment and consent.

Because of the nature of the coaching app intervention, participants and intervention providers could not be blinded to group assignment. To minimize the potential influence of group allocation on baseline assessment, group assignment was disclosed by an intervention provider who was not involved in participant recruitment or enrollment to patients and family members only after the patient had completed the baseline survey. The analyst conducting the outcome analyses was not involved in participant recruitment, group allocation, or data collection and will not be blinded to group assignment during the analyses. To minimize potential analytical bias, the analyses will be conducted according to the prespecified statistical analysis methods described in this protocol.

Sample size was calculated using G*Power (version 3.1.9.7) [35], with self-management behavior as the primary outcome. A medium effect size (f=0.25) [36] was assumed based on findings from a previous RCT that reported Cohen d values of 0.45 to 0.59 [25]. For a repeated-measures ANOVA with a group-by-time interaction, 2 groups, and 2 measurement points, α=.05 and 80% power, the minimum required sample size was estimated to be 68 patients (34 per group). Allowing for an anticipated 20% attrition rate, the planned sample size was increased to 82 patients (41 per group). Under the original design, 1 family member per patient was also expected to participate. Although the sample size was based on an efficacy-oriented calculation, clinical outcomes in this pilot RCT will be interpreted as exploratory findings rather than as definitive evidence of intervention effectiveness.

Throughout the study, participants in both groups continued their usual clinical care and prescribed treatments. Any changes in medical treatment were made at the discretion of their treating physicians.

This pilot trial is considered a minimal-risk behavioral intervention. Potential risks include temporary psychological discomfort, time burden, fatigue, technical difficulties, and privacy concerns. Before implementation, intervention content and procedures were reviewed using prior research and expert content validation to minimize burden and harm.

Participants can report concerns during coaching sessions, follow-up assessments, or direct communication with the research team. The principal investigator and research team will review each concern and take appropriate action, including technical support, further assessment, or intervention interruption, modification, or discontinuation. Outcome assessments may continue after intervention discontinuation unless the participant requests withdrawal or safety considerations require otherwise. No formal adverse-event grading or causality-classification system is prespecified. Serious or medically significant events will be promptly reported to the principal investigator and institutional review board in accordance with institutional requirements. Participant-reported concerns and unintended effects will be summarized descriptively to inform intervention feasibility and acceptability.

All analyses will be performed using SPSS Statistics (version 29.0; IBM Corp). Baseline characteristics will be summarized descriptively by randomized group without statistical significance testing.

The primary exploratory analysis will compare change scores from baseline to postintervention between the intervention and control groups using independent-samples t tests or Mann-Whitney U tests, as appropriate. Between-group effect estimates will be reported with 95% CIs and effect sizes (Cohen d or r). Within-group changes will be examined as supplementary analyses using paired-samples t tests or Wilcoxon signed rank tests. An analysis of covariance will also be conducted as a supplementary analysis to compare postintervention outcomes between groups, adjusting for the corresponding baseline values. All statistical tests will be 2-sided with a significance level of P<.05. The primary analysis population will comprise participants with both baseline and postintervention data for the outcome being analyzed. Missing values will not be imputed; their extent and reported reasons will be summarized by randomized group. Recruitment, retention, intervention adherence, and data completeness will be summarized descriptively. No separate per-protocol analysis is planned.

This study was approved by the institutional review board of Yonsei University Health System, Severance Hospital (approval 4-2025-0117), covering the RCT, smartphone app development, usability evaluation, and expert content validation. The trial was conducted in accordance with the Declaration of Helsinki. Trained research staff obtained written informed consent from all patient and family participants before enrollment, after providing written study information and sufficient time for questions. Consent covered survey and EMR data use, including related ancillary analyses. Participation in all study activities, including the RCT involving patients and family members, the usability evaluation, and expert content validation, was voluntary. All participants could withdraw at any time without penalty or effect on their medical care or participation. Information collected through these study activities was deidentified, handled confidentially, and stored securely in accordance with institutional policies. Although not prespecified in the study protocol, periodic independent reviews were conducted to confirm that no ineligible participants had been enrolled and that informed consent documentation and study procedures were appropriately followed in accordance with the approved protocol. Participants who completed the 12-week study received compensation, and those who completed the usability evaluation received separate compensation.

The findings of this pilot RCT will be reported in accordance with the CONSORT-eHEALTH (Consolidated Standards of Reporting Trials of Electronic and Mobile Health Applications and Online Telehealth) checklist [49]. Results will be disseminated through publication in peer-reviewed journals and presentation at academic conferences. No individually identifiable information will be disclosed.

This pilot RCT is expected to provide preliminary evidence regarding the feasibility of an IMB model–based mHealth coaching intervention designed to incorporate family support for middle-aged men with CLD and metabolic risk factors. The findings will help identify practical considerations related to recruitment, retention, intervention adherence, data completeness, and the delivery of family-supported digital coaching. Preliminary changes in patient- and family-related outcomes will be interpreted as exploratory and used to refine the intervention and trial procedures for a future definitive RCT.

The IMB model provides an established theoretical foundation for behavior change interventions in chronic disease management. Theory-informed interventions have supported adherence and health behaviors in DM [50,51], chronic obstructive pulmonary disease [52], and coronary artery disease [53]. In CLD, limited disease-related knowledge [54-56] and difficulty sustaining motivation for lifestyle modification [57] remain important self-management challenges. Inadequate engagement is also a persistent limitation of smartphone-based health interventions [58]. By addressing information, personal motivation, social motivation, and behavioral skills within a structured digital environment, this study may provide useful insights into the feasibility of applying the IMB model to family-supported CLD self-management.

A key contribution of this study is the integration of structured family participation into an mHealth self-management intervention. The intervention was designed to extend individual-centered care by embedding social motivation and opportunities for family support within a technology-supported environment. Prior research suggests that family involvement may facilitate lifestyle modification [59], improve psychological outcomes such as reduced anxiety and greater self-efficacy [60], and support self-management in mHealth interventions [61]. Despite this growing body of evidence [62-65], family members have often been involved only peripherally, and sustaining their engagement over time remains challenging [66-68]. By incorporating family participation as a core intervention component rather than an adjunct [69,70], this pilot RCT will provide preliminary evidence on the feasibility of delivering a family-supported mHealth self-management intervention and inform strategies for strengthening family engagement in a future definitive trial.

This study has some limitations. First, the feasibility and generalizability of this mHealth-based intervention may be limited among individuals with low digital literacy or restricted access to mobile devices. Second, recruitment of patient-family dyads remained challenging. Enrollment was limited when family members lived separately in another region, were substantially older or younger than the patient, or were unable to use a mobile phone. As a result, the planned sample size was not reached. The reduced sample may have limited statistical power and the precision of effect estimates; therefore, the clinical outcomes will be interpreted as exploratory. Third, the usability evaluation involved a small number of participants assessed at different stages of app use, which may limit the generalizability of the usability findings.

This protocol describes a single-center pilot RCT of an IMB model–based, family-supported mHealth coaching intervention for middle-aged men with CLD and metabolic risk factors. The intervention integrates app-based self-monitoring, individualized coaching, and family participation to support sustained self-management. Of the target enrollment of 82 patients, 52 were enrolled, and 35 completed the 12-week follow-up. Given the smaller-than-planned sample, clinical outcomes will be interpreted as exploratory rather than as definitive evidence of intervention effectiveness. The findings will inform refinement of the intervention, recruitment and retention strategies, and the design of a future adequately powered RCT. Following completion of the ongoing analysis, the findings will be disseminated through peer-reviewed publication and academic presentations.