Skip to main content
OpenTrials
Completed

NCT Number: NCT07455916

Short-Term Effects of an AI-Based Wearable Adherence Monitor in Outpatient Psychiatry

Medication nonadherence undermines treatment effectiveness in psychiatric care, yet objective measurement in routine practice remains challenging. AI-enabled wearables may offer scalable monitoring, but evidence from randomized evaluations is limited. This exploratory trial evaluated the short-term effect of an AI-enabled smartwatch intervention on clinician-assessed medication adherence in adolescent and young adult psychiatric outpatients.

Completed

Looking for future studies?

Notify Me

Key information

Conditions

Mental Disease Adherence to Care Behavior Digital Health Health Behavior Medication Adherence Mental Disorders Patient Acceptance of Health Care Patient Compliance Treatment Adherence and Compliance

Age range

12 year and older

Sex eligibility

All sexes

Study type

Interventional

Phase

Not applicable

Primary location

Wonkwang University Hospital

Iksan, Jeollabuk-do, 54538, South Korea

About this study

We conducted a single-center, prospective, exploratory randomized controlled trial of an AI-enabled smartwatch adherence-monitoring system in outpatient psychiatric care. Participants aged ≥12 years were randomized 1:1 to smartwatch intervention or usual care for 4 weeks. The primary outcome was the between-group difference in change in Brief Adherence Rating Scale (BARS) score (Δ = post - pre) over prespecified 28-day baseline and post-intervention windows. The BARS is a clinician-administered instrument that incorporates pill-count components (prescribed count, actual count taken, missed-dose frequency) integrated into a 0-100% visual analog score. Prespecified secondary analyses included responder thresholds at Δ ≥10 and ≥20 percentage points. Exploratory analyses examined the association between device engagement and BARS change score, and a within-subject comparison in waitlist participants who subsequently completed the smartwatch period. Analyses used complete cases with linear regression (adjusting for baseline adherence, age, and sex) and HC3 robust standard errors; ANCOVA served as a confirmatory model. Pre-specified sensitivity analyses included bootstrap resampling, Winsorized regression, and Wilcoxon rank-sum testing. Because of the exploratory/feasibility design, the a priori sample size calculation used α = 0.20; all inferential analyses were reported at two-sided α = 0.05.

Who can participate

Healthy volunteers accepted: No

Only the study team can determine whether someone qualifies for participation.

Inclusion criteria

  • Age 12 years or older
  • Receiving outpatient psychiatric care at the Department of Psychiatry
  • Willing to wear a study smartwatch and undergo study assessments
  • Able to provide written informed consent (or assent with parental consent for minors)

Exclusion criteria

  • Age under 12 years
  • Diagnosed with congenital genetic disorders
  • Clear history of acquired brain injury (e.g., cerebral palsy, cerebral infarction)
  • Comorbid seizure disorders or other neurological conditions
  • Severe psychotic symptoms or aggression precluding safe and stable smartwatch use or posing risk of device damage

Treatment and study plan

AI-enabled smartwatch

Device

The device was designed to monitor medication-related behaviors in real-world settings (pill taking and, by design, use of eye drops, inhalers, and nasal sprays). A built-in camera remained in sleep mode and recorded brief ~20-second clips only when an electronic tag affixed to the medication container signaled three concurrent conditions: (1) container motion detected by the tag's accelerometer, (2) ambient light detected by the tag's light sensor, and (3) watch-tag proximity within approximately 10-15 cm via BLE ranging. After capturing a clip, the camera returned to sleep. Encrypted videos were transmitted to a secure server and linked to de-identified study IDs. Server-side algorithms then analyzed the full 20-second sequence, covering the continuous hand actions from opening to closing of the container, and returned a binary medication event (medication vs no medication). Participants were instructed to wear the smartwatch for 4 weeks.

Primary outcomes

  1. Change in Brief Adherence Rating Scale (BARS) Score

    Time frame: Baseline (pre-randomization 28-day window) and end of intervention (28-day post-intervention window, week 4)

    Medication adherence was assessed using the Brief Adherence Rating Scale (BARS), a clinician-administered instrument that structurally incorporates pill-count components (Byerly et al., 2008). At each assessment, the clinician elicits the number of pills prescribed over the preceding 28 days, the number actually taken, and the frequency of missed doses, and integrates these inputs with the structured patient interview into a 0-100% visual analog score. The primary outcome is the between-group difference in BARS change score (Δ = post - pre).

Secondary outcomes

  1. Responder Rate at Δ ≥ 10 Percentage Points

    Time frame: Baseline to end of intervention (4 weeks)

    Proportion of participants achieving a BARS change score of at least 10 percentage points from baseline. Between-group comparison using logistic regression to estimate odds ratio and 95% confidence interval.

  2. Responder Rate at Δ ≥ 20 Percentage Points

    Time frame: Baseline to end of intervention (4 weeks)

    Proportion of participants achieving a BARS change score of at least 20 percentage points from baseline. Between-group comparison using logistic regression to estimate odds ratio and 95% confidence interval.

Other outcomes

  1. Association Between Device Engagement and BARS Change Score

    Time frame: 4 weeks (intervention perio

    Among intervention-arm participants, association between device engagement rate (proportion of intervention days with ≥1 device-recorded medication event) and BARS change score, analyzed using Pearson correlation and linear regression.

  2. Within-Subject Adherence Change in Waitlist Crossover Cohort

    Time frame: Baseline, end of waitlist observation (week 4), and end of smartwatch period (week 8)

    Among control-arm participants who subsequently completed the smartwatch period, comparison of adherence measured during the waitlist observation phase versus the subsequent smartwatch period. Paired comparison using t-test.

Sponsors and collaborators

Lead sponsor

Wonkwang University Hospital

Other

Registry information

Official study title

Short-Term Effects of an AI-Based Wearable Adherence Monitor in Outpatient Psychiatry: Randomized Controlled Trial.

Important dates

Study start
2022
Primary completion
2025
Study completion
2025
First posted
Mar 6, 2026
Registry last updated
Apr 23, 2026

OpenTrials presents study information sourced from ClinicalTrials.gov. The official registry record should be consulted for the latest information.

View the official ClinicalTrials.gov record (opens in a new tab)

This listing is for discovery and informational purposes only. It is not medical advice, does not guarantee that a study is recruiting, and does not determine eligibility. Contact the study team and a qualified healthcare professional when considering participation.

Related clinical trials

Published trials that share one or more normalized conditions with this study.