Predicting adherence to fully-automated, chatbot-delivered digital cognitive behavioral therapy for insomnia (dCBT-I) using machine learning: A pilot real-world study
Vorhersage der Adhärenz zu vollständig automatisierter, chatbot-gestützter digitaler kognitiver Verhaltenstherapie bei Insomnie (dCBT-I) mithilfe von Machine Learning: eine Pilot-Real-World-Studie
PLOS Digital Health · 02.01.2026 · 5 (1) · e0001170
DOI: 10.1371/journal.pdig.0001170 ↗Auf einen Blick
Bei 75 Personen mit Insomnie-Symptomen sagte Machine Learning die Adhärenz zur 28-tägigen dCBT-I teils vorher. Baseline-depressive Symptome waren der stärkste Prädiktor für Nicht-Adhärenz.
Fragestellung
Lassen sich Adhärenz und Folgeadhärenz zu dCBT-I anhand von Basismerkmalen mit Machine Learning vorhersagen?
Methode
Pilot-Real-World-Studie mit 75 Personen; 28-tägiges chatbot-gestütztes dCBT-I-Programm; ML-Modelle mit Kreuzvalidierung; Datenimbalance im Training mit SMOTE behandelt.
Ergebnisse
Baseline-depressive Symptome sagten Nicht-Adhärenz am besten voraus. Höhere depressive Symptome waren mit kürzerer Nutzungsdauer verbunden (β = -3,57; 95 %-KI -5,82 bis -1,33; p = ,002). AUCs 0,65-0,91 für Gesamtadhärenz und 0,56-0,74 für Folgeadhärenz.
Schlussfolgerung der Autor:innen
Die Studie zeigt das Potenzial von ML zur Vorhersage der Nutzeradhärenz bei dCBT-I und liefert klinische Hinweise zur Personalisierung schlafbezogener Behandlungen.
Kernaussagen
- 75 Personen mit Insomnie-Symptomen nahmen an einem 28-tägigen dCBT-I-Programm teil
- Depressive Symptome waren der stärkste Prädiktor für Nicht-Adhärenz
- ML-Modelle erreichten AUCs von 0,65-0,91 für Gesamtadhärenz
Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.
Kurz gefasst
Fachartikel in der Zeitschrift PLOS Digital Health, online erschienen am 02.01.2026. Verfasst von Rose Wing Lai So und 8 weiteren Autor:innen. Beteiligte Einrichtung: Li Chiu Kong Family Sleep Assessment Unit, Department of Psychiatry, The Chinese University of Hong Kong, Hong Kong, China. Die Arbeit nennt 26 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: General Research Fund, Innovation and Technology Support Programme, Health and Medical Research Fund.
Abstract (Originalsprache)
Digital cognitive behavioral therapy for insomnia (dCBT-I) is effective in treating insomnia, but adherence remains a major challenge in real-world applications. Machine learning (ML) offers potential in predicting healthcare utilization. This study applied ML techniques to predict adherence to dCBT-I based on participant baseline characteristics. This pilot real-world study included 75 individuals (69% female; 41% aged 35–44 years) with insomnia symptoms (Insomnia Severity Index, ISI ≥ 8) who participated in a 28-day chatbot-delivered dCBT-I program. ML models, including logistic regression with elastic-net penalty, support vector machine, random forest, and gradient boosting, analyzed participant baseline characteristics to predict adherence to dCBT-I in terms of session completion, usage duration, and response volume. These models were fine-tuned using grid search and evaluated with cross-validation. The synthetic minority over-sampling technique was applied to address data imbalances in the training set. Baseline depressive symptoms were the most predictive of non-adherence. Higher depressive symptoms were associated with shorter overall usage duration (β = -3.57, 95% CI: -5.82 to -1.33, p = .002). Longer sleep onset latency and wake time after sleep onset from the previous night increased the number of responses and longer usage duration on the following day (β = 0.01-0.05, p < .05). No significant associations were found between daytime and bedtime usage and sleep parameters for that specific night. ML models predicted overall adherence, with AUCs of 0.65-0.91 (p < .05). ML models also predicted next-day adherence, with AUCs of 0.56-0.74 (p < .05). This real-world study demonstrates the potential of ML to predict user adherence to dCBT-I and provides clinical insights for personalizing sleep-focused treatments. The study also investigated daily usage and adherence patterns in dCBT-I to predict next-day adherence.
Autor:innen und Institutionen
- Li Chiu Kong Family Sleep Assessment Unit, Department of Psychiatry, The Chinese University of Hong Kong, Hong Kong, China.
- Department of Psychology, The University of Hong Kong, Hong Kong, China.
- The State Key Laboratory of Brain and Cognitive Sciences, The University of Hong Kong, Hong Kong, China.
- Li Ka Shing Institute of Health Sciences, Faculty of Medicine, The Chinese University of Hong Kong, Hong Kong, China.
ORCID-Profile: Kit Ying Chan, Yun Kwok Wing, Tim Man Ho Li
Förderung
- General Research Fund
- Innovation and Technology Support Programme
- Health and Medical Research Fund
- Knowledge Transfer Project Fund
- CUHK Direct Grant for Research
Zitieren
APA
So, R. W. L., Chan, K. Y., Cheng, C. C. W., Chan, N. Y., Li, S. X., Chan, J. W. Y., Chau, S. W. H., Wing, Y. K., & Li, T. M. H. (2026). Predicting adherence to fully-automated, chatbot-delivered digital cognitive behavioral therapy for insomnia (dCBT-I) using machine learning: A pilot real-world study. PLOS Digital Health, 5(1), e0001170. https://doi.org/10.1371/journal.pdig.0001170
Vancouver
So RWL, Chan KY, Cheng CCW, Chan NY, Li SX, Chan JWY, et al. Predicting adherence to fully-automated, chatbot-delivered digital cognitive behavioral therapy for insomnia (dCBT-I) using machine learning: A pilot real-world study. PLOS Digital Health. 2026;5(1):e0001170. doi:10.1371/journal.pdig.0001170
BibTeX
@article{So2026,
author = {So, Rose Wing Lai and Chan, Kit Ying and Cheng, Christopher Chi Wai and Chan, Ngan Yin and Li, Shirley Xin and Chan, Joey Wing Yan and Chau, Steven Wai Ho and Wing, Yun Kwok and Li, Tim Man Ho},
title = {Predicting adherence to fully-automated, chatbot-delivered digital cognitive behavioral therapy for insomnia (dCBT-I) using machine learning: A pilot real-world study},
journal = {PLOS Digital Health},
year = {2026},
volume = {5},
number = {1},
pages = {e0001170},
doi = {10.1371/journal.pdig.0001170}
}
EUVIMED