Feasibility and Preliminary Effectiveness of a Prototype Digital Therapeutic for Insomnia (LUIT-K): A Pilot Study
Article information
Abstract
Objective
Insomnia is a prevalent and burdensome condition, with limited access to evidence-based treatments like cognitive behavioral therapy for insomnia (CBT-I). Digital therapeutics offer a promising solution to bridge this treatment gap, necessitating early-phase evaluations for usability and clinical effectiveness. This pilot study aimed to evaluate the feasibility, preliminary effectiveness, and user experience of LUIT-K, a prototype mobile-based digital therapeutic delivering CBT-I content, in a Korean adult population with insomnia symptoms.
Methods
Ten adults experiencing insomnia symptoms (Insomnia Severity Index [ISI] ≥8) completed a 6-week single-arm intervention using the LUIT-K mobile app. Validated self-report measures were administered at baseline and endpoint, including the ISI, Pittsburgh Sleep Quality Index-Korean version (PSQI-K), Korean Epworth Sleepiness Scale (KESS), Patient Health Questionnaire-9 (PHQ-9), Generalized Anxiety Disorder-7 (GAD-7), Dysfunctional Beliefs about Sleep (DBAS), Sleep Hygiene Practices Scale (SHPS), and Korean version of the Biological Rhythms Interview of Assessment in Neuropsychiatry (K-BRIAN). Usability and user experience were assessed using an 11-item structured survey and open-ended feedback. Responder analysis was conducted based on a ≥6-point improvement in ISI scores.
Results
Significant reductions were observed in ISI scores, decreasing from a mean of 16.7 (SD=5.1) at baseline to 11.9 (SD=5.0) at endpoint (p=0.04). This change indicates a modest clinical improvement at the group level; at the individual level, 40% of participants were classified as responders based on the ISI criterion of a 6-point reduction. Additionally, statistically significant improvements were noted in PSQI-K (p=0.02) and SHPS (p=0.01) scores, with a trend toward improvement in KESS (p=0.06) but no significant change in DBAS (p=0.38) and GAD-7 (p=0.48). Qualitative feedback highlighted technical issues and usability as the most frequent complaints.
Conclusion
This pilot study provides preliminary support for the feasibility and clinical potential of the LUIT-K digital CBT-I app. Findings suggest that early symptom improvement is associated with a more favorable user experience, underscoring the importance of iterative design informed by both clinical and user-centered data. Further large-scale controlled studies are warranted to confirm efficacy and optimize deployment strategies.
INTRODUCTION
Insomnia is one of the most prevalent and burdensome sleep disorders, affecting approximately 10%–30% of the global adult population, depending on diagnostic criteria and region [1]. In Korea, the prevalence of chronic insomnia has increased steadily over the past decade, with a notable rise in sleep-related complaints among young and middle-aged adults [2]. Insomnia is associated with a range of negative outcomes, including impaired cognitive and occupational functioning, increased risk of psychiatric disorders such as depression and anxiety, and a higher incidence of cardiometabolic disease [3].
Cognitive behavioral therapy for insomnia (CBT-I) is the recommended first-line treatment according to clinical guidelines, with robust evidence supporting its efficacy in both short- and long-term outcomes [4,5]. However, despite its effectiveness, access to CBT-I remains limited due to a shortage of trained providers, logistical barriers, and the time-intensive nature of delivery formats. As a result, many individuals with insomnia rely on pharmacologic interventions or do not receive any treatment at all [1].
Recent advances in digital health technology have led to the development of digital therapeutics that deliver CBT-I content via smartphone apps or web-based platforms [6,7]. Digital CBT-I interventions have shown promise in improving sleep outcomes, enhancing scalability, and increasing patient engagement [8]. Meta-analyses of randomized trials suggest that app-based CBT-I can be as effective as traditional face-to-face therapy in reducing insomnia symptoms [9]. Nonetheless, further research is needed to evaluate the feasibility, acceptability, and real-world effectiveness of digital CBT-I tools, particularly during early stages of development.
In addition to clinical efficacy, understanding user experience and treatment engagement is essential for the successful implementation of digital health interventions [6]. As digital solutions are inherently self-guided and often lack interpersonal support, factors such as perceived usability, credibility, and personal relevance play a critical role in treatment outcomes. Identifying patterns of symptom response and usability feedback can inform design refinements and improve therapeutic impact.
The present study aimed to conduct a pilot feasibility trial of a prototype digital CBT-I application developed for Korean users. We evaluated 1) changes in insomnia and related symptoms over a 6-week intervention period, 2) symptom improvement patterns among treatment responders vs non-responders, and 3) user-reported experience and feedback regarding the app’s content and usability. The study also included open-ended feedback to explore user-perceived barriers and suggestions for future improvement. As an exploratory investigation, the goal was not to test specific hypotheses but to generate insights for further clinical validation and product development.
METHODS
Study design and setting
This study was a single-arm, prospective pilot trial conducted over a 6-week period to evaluate the feasibility, preliminary effectiveness, and user experience of a digital therapeutic prototype targeting insomnia symptoms. The intervention was delivered via a mobile application designed based on CBT-I, named LUIT-K (LumanLab Inc., Sejong, Korea). The study was approved by the Institutional Review Board (IRB) of Korea University Anam Hospital (IRB No. 2024AN0538). All participants provided written informed consent prior to enrollment.
Participants
Participants were recruited from the outpatient psychiatric and sleep clinics at Korea University Anam Hospital. Eligibility criteria included adults aged 19 to 70 years who had experienced insomnia symptoms at least three times per week for the past three months, and scored ≥8 on the Insomnia Severity Index (ISI). Individuals with a diagnosis of psychotic spectrum disorders, significant neurological conditions, or other primary sleep disorders (e.g., narcolepsy, obstructive sleep apnea) were excluded. Participants were also required to own a smartphone to access the intervention. A total of 10 participants were enrolled in this pilot study, which aimed to investigate feasibility.
Intervention
The intervention consisted of a 6-week program delivered exclusively through a mobile application (LUIT-K) developed to provide digital CBT-I content (Figure 1). The app consisted of six weekly modules that covered the major components of CBT-I. In week 1, users were introduced to the principles of normal sleep and completed an initial digital sleep diary. Week 2 addressed stimulus control techniques, while week 3 focused on sleep restriction. In week 4, sleep hygiene education was provided. Week 5 introduced cognitive restructuring strategies to address maladaptive sleep-related beliefs. Finally, week 6 included training in relaxation techniques such as diaphragmatic breathing and progressive muscle relaxation.
Screenshots of prototype digital therapeutic for insomnia (LUIT-K, LumanLab). Images are reproduced with permission from LumanLab Inc.
Each weekly module incorporated psychoeducational videos, interactive exercises, quizzes, and self-monitoring tasks. Participants were also encouraged to log their daily sleep behavior via an in-app sleep diary, which collected bedtime, wake time, sleep latency, nocturnal awakenings, and subjective sleep quality. The app provided automated weekly feedback on sleep efficiency and offered gamified rewards (e.g., virtual coins and an animated “apple tree”) to encourage engagement.
Measures
Participants completed a battery of self-report questionnaires at baseline and at the 6-week endpoint. The primary outcome was the ISI [10], a 7-item scale measuring the severity and impact of insomnia symptoms over the past two weeks. Each item is rated on a 5-point Likert scale (0–4), yielding a total score range of 0 to 28. Higher scores indicate greater severity of insomnia. The Pittsburgh Sleep Quality Index-Korean version (PSQI-K) was administered to assess global sleep quality across seven domains, including sleep latency, duration, efficiency, and disturbances [11]. Total scores range from 0 to 21, with higher scores indicating poorer sleep quality. To assess excessive daytime sleepiness, participants completed the Korean Epworth Sleepiness Scale (KESS) [12], which includes 8 items rated on a 0–3 scale.
The Generalized Anxiety Disorder-7 (GAD-7) scale was used to measure the severity of anxiety symptoms [13]. Each of the 7 items is scored from 0 (not at all) to 3 (nearly every day), with total scores ranging from 0 to 21. Depressive symptoms were assessed with the Patient Health Questionnaire-9 (PHQ-9), a widely used 9-item instrument for detecting and quantifying depression [14]. Each item is scored 0–3, with total scores ranging from 0 to 27.
The Dysfunctional Beliefs and Attitudes about Sleep (DBAS) scale was included to evaluate maladaptive cognitive patterns related to sleep [15]. It consists of 16 items, each rated on an 11-point scale from 0 (strongly disagree) to 10 (strongly agree). Higher scores reflect stronger dysfunctional beliefs. Sleep hygiene behaviors were measured using the Sleep Hygiene Practice Scale (SHPS), which includes 30 items assessing sleep-related routines, behaviors, and environmental factors [16]. Items are scored from 1 to 6, with higher scores indicating poorer sleep hygiene.
To evaluate circadian rhythm and biological rhythm stability, participants completed the Korean version of the Biological Rhythms Interview of Assessment in Neuropsychiatry (K-BRIAN) [17]. The scale includes 21 items covering domains such as sleep-wake timing, activity levels, eating patterns, and social rhythms. Each item is rated on a 1–4 scale, with higher scores indicating more irregular biological rhythms.
To explore treatment responsiveness, participants were categorized into responder and non-responder groups based on changes in ISI scores. Responders were defined as those who showed a reduction of 6 or more points on the ISI between baseline and endpoint, a threshold consistent with previously established minimally clinically important difference values for clinical insomnia [17].
Usability and user experience evaluation
To assess the usability and user experience of the digital therapeutic software prototype for insomnia, a comprehensive questionnaire was developed and administered to participants. This questionnaire aimed to capture users’ perceptions across several key dimensions, including the software’s ease of use, its impact on understanding and managing insomnia, perceived treatment effectiveness, and future intent to use. Participants were asked to rate their agreement with statements regarding various aspects of their experience, alongside providing feedback on areas for improvement. Specifically, the evaluation included 11 items: 9 statements assessed on a Likert scale (e.g., “The software was easy to use,” “My insomnia symptoms improved after using the software”), and two open-ended questions designed to identify the primary pain points and areas for improvement from the user’s perspective (e.g., “What was your biggest complaint while using the software?” and “What area most needs improvement in the software prototype?”). The insights gathered from this evaluation are crucial for refining the software and enhancing its efficacy and user satisfaction.
Statistical analysis
Descriptive statistics were computed for demographic and outcome variables. Changes in ISI and other symptom scales were analyzed using the Wilcoxon signed-rank test, appropriate for small, non-normally distributed samples. To ensure transparency and consistency with our non-parametric analysis, median and interquartile range (IQR) values for all scales at baseline, endpoint, and change are provided in Supplementary Table 1. Differences between responder and non-responder groups were evaluated using the Mann–Whitney U test. Responses to user feedback items were analyzed descriptively, and distributions were visualized using boxplots. All statistical analyses and visualizations were performed using Python 3.13.3 (pandas, scipy, seaborn, matplotlib; https://www.python.org/).
RESULTS
Participant characteristics
A total of 10 participants completed the 6-week intervention. The average age was 28.8 years (SD=2.4), with 7 participants identifying as female and 3 as male. Most participants reported regular daytime activity patterns, did not engage in shift work, and had stable sleep environments. Educational backgrounds were predominantly university-level.
Changes in insomnia, sleep-related symptoms, and emotional symptoms
The primary outcome of this pilot study was a change in insomnia severity, as measured by the ISI. From baseline to endpoint, the mean ISI score decreased significantly from 16.7 (SD= 5.12) to 11.9 (SD=5.02), resulting in an average reduction of 4.8 points (SD=6.53). This reduction was statistically significant (Wilcoxon signed-rank test, p=0.04). Forty percent of participants were classified as responders based on the ISI criterion.
Sleep quality also improved during the study. The PSQI-K score decreased from a baseline of 20.2 to 17.0 at endpoint (Δ= −3.2, p=0.02), suggesting modest but statistically significant improvements in subjective sleep quality. The high baseline PSQI-K score of 20.2, near the upper bound of 21, is consistent with a sample reporting severe subjective sleep disturbance and suggests the possibility of a ceiling effect. Improvements were also seen in sleep hygiene, with SHPS scores dropping from 93.5 (SD=15.99) to 80.8 (SD=20.69) (Δ=−12.7, p=0.01). These results indicate that the intervention not only improved insomnia symptoms but also had a positive impact on sleep hygiene and daily functioning. Additionally, statistically significant improvements were noted in PSQI-K (p=0.02) and SHPS (p=0.01) scores, with a trend toward improvement in KESS (p=0.06) but no significant changes in DBAS (p=0.38) and GAD-7 (p=0.48). The results are shown in Table 1 and Figure 2.
Baseline and endpoint scores and mean changes for insomnia, sleep-related, and emotional symptoms during digital therapeutic intervention
Mean changes in scale scores from baseline to endpoint following digital therapeutic intervention. ISI, Insomnia Severity Index; PHQ-9, Patient Health Questionnaire-9; GAD-7, Generalized Anxiety Disorder-7; KESS, Korean Epworth Sleepiness Scale; DBAS, Dysfunctional Beliefs and Attitudes about Sleep; SHPS, Sleep Hygiene Practice Scale; K-BRIAN, Korean version of the Biological Rhythms Interview of Assessment in Neuropsychiatry; PSQI-K, Pittsburgh Sleep Quality Index-Korean version.
Responder vs. non-responder group comparisons
To explore response heterogeneity, participants were classified into responder (n=4) and non-responder (n=6) groups based on a ≥ 6-point decrease in ISI scores. Across all measured outcomes, responders exhibited greater improvement than non-responders. For example, responder participants showed larger decreases in GAD-7 (−3.0 vs. 0.0) and PHQ-9 (−2.0 vs. −0.8). Similarly, they reported stronger improvements in sleep-related beliefs (DBAS: −19.3 vs. −1.7) and sleep hygiene behaviors (SHPS: −21.5 vs. −6.8). Although these group differences were not statistically significant due to small sample size, they suggest a consistent pattern of greater overall benefit among responders.
User experience and app evaluation
Participants provided usability feedback via an 11-item survey administered at the endpoint. Among them, 9 items reported overall ratings of software effectiveness (Figure 3). Participants also responded to two open-ended survey questions: one regarding dissatisfaction with the app (Q10) and the other asking which aspects should be improved (Q11). Across all participants, “technical issues” and “usability” were the most frequently mentioned dissatisfaction themes.
Distribution of usability and user experience scores at the endpoint.
DISCUSSION
This pilot study examined the feasibility and clinical utility of a prototype digital CBT-I application among adults experiencing insomnia symptoms. Over a 6-week intervention period, participants demonstrated statistically and clinically meaningful reductions in insomnia severity, along with improvements in sleep quality, dysfunctional sleep beliefs, and anxiety symptoms. Furthermore, responder group analysis revealed consistent trends indicating that individuals who experienced greater clinical benefit also reported more favorable user experiences and higher app engagement.
One of the most notable findings was the significant decrease in ISI scores, with an average reduction of 4.8 points. This magnitude of change aligns with prior pilot studies of digital CBT-I, where ISI improvements in the range of 4–8 points were observed after similar durations [9]. Although not all participants met the conventional 6-point threshold for “response,” the overall improvement supports the potential effectiveness of the digital prototype in a real-world outpatient setting. In addition to insomnia symptoms, participants showed statistically significant improvements in sleep hygiene (SHPS) and subjective sleep quality (PSQI-K). The improvement in sleep hygiene is particularly encouraging, given that these changes are essential to sleep health [18].
Emotional symptoms, including anxiety and depression, also trended toward improvement [19]. The significant decrease in GAD-7 scores suggests that better sleep may contribute to emotional regulation, consistent with prior research indicating that insomnia treatment can yield downstream effects on mood. However, the modest reduction in PHQ-9 scores did not reach statistical significance, possibly due to limited sample size or floor effects among participants with mild baseline symptoms.
Analysis of responder and non-responder subgroups provided valuable insight into treatment heterogeneity. Responders demonstrated more pronounced improvements across sleep, cognitive, and emotional domains, and rated the app more favorably on usability and effectiveness dimensions. While these trends did not reach statistical significance, they offer practical implications for identifying early markers of treatment success.
In terms of usability, participants generally found the app easy to use and helpful. Ratings for treatment expectations, perceived benefit, and future willingness to use the full version of the app increased over time, particularly among responders. These results underscore the importance of evaluating not only clinical outcomes but also perceived value from the user’s perspective [20]. The open-ended feedback further contextualized these findings. The most commonly cited areas for dissatisfaction were technical issues and navigation difficulties. This pattern echoes prior usability studies indicating that negative user experience may compound treatment non-response.
This study has several strengths. First, it used both quantitative and qualitative data to comprehensively assess feasibility, symptom change, and user feedback. Second, the responder analysis allowed exploration of outcome variability and highlighted the value of individual-level insights. Third, the digital therapeutic was tested in a naturalistic, self-guided format, enhancing ecological validity.
However, limitations must be acknowledged. As a pilot study with only 10 participants and no control group, statistical power was limited, and generalizability is constrained. The short intervention period (6 weeks) may have been insufficient to observe the full therapeutic effects in some users. Additionally, app usage data (e.g., adherence, module completion) was not systematically tracked and could not be linked to outcomes. Finally, the use of self-report measures may introduce bias, and future studies should consider incorporating objective sleep metrics.
To build on these findings, future research should involve a larger sample size and a randomized controlled design. The inclusion of app usage metrics and real-time adherence tracking could shed light on the mechanisms linking engagement and outcomes. Furthermore, integrating adaptive features such as tailored content or human guidance may enhance personalization and effectiveness. Finally, longer-term follow-up is necessary to assess the maintenance of gains and prevent relapse.
In conclusion, this pilot study provides preliminary evidence that a mobile-based digital CBT-I prototype is feasible, usable, and may contribute to meaningful reductions in insomnia and related symptoms. Participants who benefited clinically also tended to report higher satisfaction and engagement with the intervention. These results underscore the potential of digital therapeutics to bridge the accessibility gap in behavioral sleep treatment. While further validation through controlled trials is warranted, the findings offer valuable insight for developers, clinicians, and implementation scientists seeking to optimize next-generation digital mental health tools.
Supplementary Materials
The Supplement is available with this article at https://doi.org/10.33069/cim.2025.0041.
Median and interquartile range (IQR) of scores at baseline and endpoin
Notes
Chul-Hyun Cho, a contributing editor of Chronobiology in Medicine, was not involved in the editorial evaluation or decision to publish this article. The remaining authors have no potential conflicts of interest to disclose.
Availability of Data and Material
The datasets generated or analyzed during the study are available from the corresponding author on reasonable request.
Author Contributions
Conceptualization: Solee Han, Chul-Hyun Cho. Data curation: all authors. Formal analysis: all authors. Funding acquisition: Chul-Hyun Cho. Investigation: all authors. Methodology: all authors. Project administration: Chul-Hyun Cho. Resources: Seung Pil Pack, Chul-Hyun Cho. Software: Seung Pil Pack. Supervision: Chul-Hyun Cho. Validation: all authors. Visualization: Solee Han, Seung Pil Pack. Writing—original draft: Solee Han. Writing—review&editing: all authors.
Funding Statement
This work was supported by the National Research Foundation of Korea (NRF) grants (NRF-2021R1A5A8032895 and NRF-2022M3C1B6080866), a grant of the Institute of Information & communications Technology Planning & Evaluation (IITP) (RS-2023-00224823), and a grant of the Information and Communications Promotion Fund through the National IT Industry Promotion Agency (NIPA) (H0601-24-1017), funded by the Ministry of Science and Information and Communications Technology (MSIT), Republic of Korea. The sponsors had no role in study design; in the collection, analysis, and interpretation of data; in the writing of the report; and in the decision to submit the paper for publication.
Acknowledgments
We would like to thank LumanLab Inc. for their collaboration in app development and implementation throughout the research.