Temporal Causal Model for Medication Adherence Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for evaluating medication adherence in patients are inadequate in identifying personalized risk factors and providing timely interventions, leading to unclear strategies for preventing non-adherence, particularly in chronic conditions like Type 2 diabetes mellitus.
Innovation Solution
A computer-based system that uses temporal causal models to predict medication adherence by identifying risk factors and determining personalized interventions, incorporating patient clusters, causality measures, and dynamic features to assess adherence behavior and recommend targeted interventions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If current methods for evaluating medication adherence are used, then general adherence assessment is possible, but personalized risk factor identification and timely intervention capabilities are insufficient
Solution Approach 1:
The patent segments patients into distinct clusters based on adherence patterns and risk factors. By dividing the patient population into homogeneous groups with similar characteristics, the system can identify personalized risk factors for each cluster and apply targeted interventions, thereby improving both measurement precision and adaptability simultaneously
Solution Approach 2:
The patent applies local quality by providing different intervention strategies tailored to specific patient clusters rather than uniform treatment. Each cluster receives customized adherence promotion plans based on their unique risk factor profiles, enabling precise personalization while maintaining overall system effectiveness
2Measurement precision
If temporal causal models with multiple features are implemented, then personalized prediction accuracy improves, but system complexity increases
Solution Approach 1:
The patent reduces model complexity by segmenting patients into clusters first, then building simplified temporal causal models for each cluster rather than creating one complex model for the entire population. This segmentation allows the use of fewer features per cluster while maintaining high prediction accuracy
Solution Approach 2:
The patent changes parameters by selecting different subsets of features for different patient clusters based on their specific characteristics. Rather than using all possible features uniformly, the system adapts the feature set to each cluster's needs, reducing overall complexity while preserving predictive power
Data Source
AI summary
A method, a computer program product, and a computer system predict medication adherence of a patient. The method includes identifying risk factors associated with medication adherence of the patient. The method includes determining a likely behaviour for medication adherence of the patient based on the identified risk factors and a temporal causal model. The temporal causal model is based on features of a patient cluster to which the patient belongs. The features are nodes in the temporal causal model. The likely behaviour is based on causality measures for each identified risk factor to the nodes. The method includes determining a current medication adherence value of the patient. The current medication adherence value is indicative of a ratio between an actual medication regiment and an expected medication regiment. The method includes determining a future medication adherence value of the patient based on the current medication adherence value and the causality measures.