Predicting Patient-Level 3-Level Version of EQ-5D Index Scores From a Large International Database Using Machine
Zsombor Zrubka1, István Csabai2, Zoltán Hermann3
1Health Economics Research Center, University Research and Innovation Center, Óbuda University, Budapest, Hungary; Corvinus Institue for Advanced Studies, Corvinus University of Budapest, Budapest, Hungary.
Summary
Machine learning and regression models were evaluated for predicting EQ-5D-3L scores. Regression methods like XGBR and OLS showed better performance, but prediction errors often exceeded clinical relevance.
Area of Science:
- Health outcomes research
- Machine learning applications in healthcare
- Health economics and outcomes research
Background:
- The EQ-5D-3L is a widely used measure of health status.
- Accurate prediction of EQ-5D-3L index scores is crucial for health economics and outcomes research.
- Evaluating machine learning and regression models can improve prediction accuracy.
Purpose of the Study:
- To evaluate the performance of machine learning (eXtreme Gradient Boosting classification and regression) and ordinary least squares regression in predicting EQ-5D-3L index scores.
- To compare prediction accuracy across different scenarios, including general population, patients, and combined samples, using various predictor sets.
- To assess model performance based on mean absolute error and classification into clinically relevant health severity groups.
Main Methods:
- Combined data from 30 studies across 3 countries, involving 26,318 individuals.
- Employed eXtreme Gradient Boosting classification (XGBC), eXtreme Gradient Boosting regression (XGBR), and ordinary least squares (OLS) regression.
- Utilized 10-fold cross-validation with an 80%/20% train-test split, evaluating 6 prediction scenarios with demographic, disease-related, and patient-reported outcome variables.
Main Results:
- Regression models (XGBR and OLS) generally outperformed XGBC in predicting EQ-5D-3L scores.
- For the total sample using all predictors, XGBR achieved the lowest mean absolute error (0.126) and highest percentage within the correct health severity group (64.9%).
- Despite model improvements, prediction errors remained outside the clinically irrelevant range for a majority of respondents, highlighting limitations in current prediction accuracy.
Conclusions:
- Regression models demonstrated superior performance over classification models for EQ-5D-3L index score prediction.
- The study underscores the critical need for systematic collection of patient-reported outcome data, such as EQ-5D.
- Collaboration between artificial intelligence experts and outcomes researchers is recommended to maximize the utility of health system data.
More Related Videos
Related Concept Videos
01:11Residuals and Least-Squares Property
7.9K
The vertical distance between the actual value of y and the estimated value of y. In other words, it measures the vertical distance between the actual data point and the predicted point on the line
If the observed data point lies above the line, the residual is positive, and the line underestimates the actual data value for y. If the observed data point lies below the line, the residual is negative, and the line overestimates the actual data value for y.
The process of fitting the best-fit...
If the observed data point lies above the line, the residual is positive, and the line underestimates the actual data value for y. If the observed data point lies below the line, the residual is negative, and the line overestimates the actual data value for y.
The process of fitting the best-fit...
7.9K
01:23Mechanistic Models: Compartment Models in Individual and Population Analysis
91
Mechanistic models are utilized in individual analysis using single-source data, but imperfections arise due to data collection errors, preventing perfect prediction of observed data. The mathematical equation involves known values (Xi), observed concentrations (Ci), measurement errors (εi), model parameters (ϕj), and the related function (ƒi) for i number of values. Different least-squares metrics quantify differences between predicted and observed values. The ordinary least...
91