Related Experiment Video
Updated: Apr 10, 2026

Digital Home-Monitoring of Patients after Kidney Transplantation: The MACCS Platform
Published on: April 12, 2021
Personalized Predictive Model to Predict Subtask Success of Medication Adherence Technologies for Older Adults With
Bincy Baby1, Ghada Elba1, Minzee Kim2
1School of Pharmacy, University of Waterloo, 10 Victoria Street South, Waterloo, ON, N2G 1C5, Canada, 1 519 888 4567 ext 21337.
This study shows that medication adherence technology (MAT) subtask success in older adults can be predicted using their characteristics. This helps in selecting appropriate technologies to reduce errors and improve adherence.
Area of Science:
- Gerontology
- Human-Computer Interaction
- Health Informatics
Background:
- Older adults face various barriers to medication management, including cognitive, physical, and environmental factors.
- Medication adherence technologies (MATs) can aid adherence, but their usability is inconsistent across users and devices.
- Existing research lacks detailed insights into feature-level usability challenges of MATs for older adults.
Purpose of the Study:
- To develop and validate a personalized predictive model for MAT subtask success in older adults.
- To account for diverse cognitive, physical, sensory, motivational, and environmental capabilities of older adults.
- To identify key predictors of successful MAT subtask completion.
Main Methods:
- A mixed-methods approach combining standardized questionnaires, cognitive walkthroughs, and semistructured interviews.
- Assessment of demographic, clinical, cognitive, physical, sensory, motivational, and environmental characteristics.
- Comparison of personalized predictive models (cosine similarity, generalized linear models) against nonpersonalized and naive models using mean square error (MSE).
Main Results:
- Both personalized and nonpersonalized models outperformed naive predictions, indicating predictability of subtask success.
- Personalized models showed optimal performance at m=0.25 during cross-validation, with slightly lower MSEs than nonpersonalized models.
- Models using performance-based vision measures (SMAT-based) consistently outperformed those using self-reported vision scores (DLTV-based).
Conclusions:
- Predicting subtask success of MATs in older adults is feasible.
- While personalization offered limited added benefit in this study, the subtask-focused approach provides valuable insights.
- This model can inform the selection of MATs, reduce usability-related errors, and enhance medication adherence outcomes.
More Related Videos
11:21Methodology for Establishing a Community-Wide Life Laboratory for Capturing Unobtrusive and Continuous Remote Activity and Health Data
Published on: July 27, 2018
12:18A Machine Learning Approach to Design an Efficient Selective Screening of Mild Cognitive Impairment
Published on: January 11, 2020
Related Concept Videos
Drug Dosing: Geriatric Patients
Pharmacokinetics in Geriatric Patients: Effect of Age on Drug Absorption
Pharmacokinetics in Geriatric Patients: Effect of Age on Drug Excretion
Drug Therapy
Antianxiety Medications
Pharmacodynamics in Geriatric Patients: Effects of Age
Pharmacokinetics in Geriatric Patients: Effect of Age on Drug Distribution