Predicting the Intention to Sign an Advance Directive: A Machine Learning Model Accounting for Cultural and System‐Level Factors
Journal of Nursing Scholarship · 2026-07 · 58 (4)
DOI: 10.1111/jnu.70105 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Nursing Scholarship, online erschienen am 04.06.2026. Verfasst von Mei-Chen Su und 4 weiteren Autor:innen. Beteiligte Einrichtung: School of Nursing, National Taipei University of Nursing and Health Sciences, Taipei City, Taiwan. Hauptthemen laut MEDLINE: Intention, Machine Learning, Advance Directives. Die Arbeit nennt 35 Literaturangaben.
Abstract (Originalsprache)
Purpose
To develop a machine learning model for predicting Taiwanese adults' intention to sign an advance directive (AD) and to identify the psychosocial, demographic, and system‐level predictors relevant to culturally sensitive nursing. This study distinguishes between the reflective process of advance care planning (ACP) and the formal legal act of AD completion, addressing the need to understand cultural and system‐level influences.
Design
This was a cross‐sectional quantitative study.
Methods
A survey was conducted with 1412 Taiwanese adults by using validated instruments, such as the Knowledge of Advance Care Planning Questionnaire and Advance Care Planning Attitude Scale. Data were analyzed using linear regression, random forest, and extreme gradient boosting models to predict the intention to sign an AD. A SHapley Additive exPlanations analysis was performed to interpret the model and investigate the effects of personal values and system‐level barriers.
Results
The extreme gradient boosting model outperformed the other models, with mean absolute error and root mean squared error values of 1.68 and 2.13, respectively. The SHapley Additive exPlanations analysis highlighted attitude toward ACP as the strongest predictor of signing intention. In addition to psychosocial factors, system‐level factors such as procedural unfamiliarity and high consultation costs emerged as key barriers. Furthermore, older age and a higher number of children were associated with a weaker intention to sign an AD, reflecting a preference for informal family consensus over formal legal documentation.
Conclusion
Machine learning models effectively identify the interplay between personal attitudes, family dynamics, and institutional conditions that shape AD‐related decision‐making. The transition from ACP dialogue to formal AD signing is determined by both cultural values and structural factors. Clinical Relevance Nurses should adopt a dual‐track strategy—supporting advance care planning through family‐inclusive dialogues and serving as “system navigators” to help patients overcome legal and financial barriers to advance directive signing. Data‐driven insights from the present study may inform precise, culturally responsive interventions that honor patient autonomy.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- School of Nursing, National Taipei University of Nursing and Health Sciences, Taipei City, Taiwan.
- Department of Nursing, Cathay General Hospital, Taipei City, Taiwan.
- Department of Nursing, Chang Gung University of Science and Technology, Taoyuan City, Taiwan.
- Graduate Institute of Artificial Intelligence and Big Data in Healthcare, Smart Healthcare Interdisciplinary College, National Taipei University of Nursing and Health Sciences, Taipei City, Taiwan.
ORCID-Profile: Mei-Chen Su, Hsiu-Chun Fang, Lee-Fen Ni, Shu-Chen Hsu, Chao-Yang Kuo
Zitieren
APA
Su, M. C., Fang, H. C., Ni, L. F., Hsu, S. C., & Kuo, C. Y. (2026). Predicting the Intention to Sign an Advance Directive: A Machine Learning Model Accounting for Cultural and System‐Level Factors. Journal of Nursing Scholarship, 58(4). https://doi.org/10.1111/jnu.70105
Vancouver
Su MC, Fang HC, Ni LF, Hsu SC, Kuo CY. Predicting the Intention to Sign an Advance Directive: A Machine Learning Model Accounting for Cultural and System‐Level Factors. Journal of Nursing Scholarship. 2026;58(4). doi:10.1111/jnu.70105
BibTeX
@article{Su2026,
author = {Su, Mei-Chen and Fang, Hsiu-Chun and Ni, Lee-Fen and Hsu, Shu-Chen and Kuo, Chao-Yang},
title = {Predicting the Intention to Sign an Advance Directive: A Machine Learning Model Accounting for Cultural and System‐Level Factors},
journal = {Journal of Nursing Scholarship},
year = {2026},
volume = {58},
number = {4},
doi = {10.1111/jnu.70105}
}
EUVIMED