Predictive modeling of rapid glaucoma progression based on systemic data from electronic medical records

Area of Science:

• Ophthalmology
• Glaucoma Research
• Medical Data Analysis

Background:

• Primary open-angle glaucoma (POAG) is a leading cause of irreversible blindness.
• Predicting the rate of retinal nerve fiber layer (RNFL) thinning is crucial for managing POAG progression.
• Systemic factors' role in POAG progression requires further elucidation.

Purpose of the Study:

• To identify baseline systemic features that predict rapid RNFL thinning in POAG patients.
• To develop and validate a predictive model for POAG progression using electronic medical records (EMRs).
• To understand the relationship between systemic health and ocular disease progression.

Main Methods:

• Retrospective analysis of EMR data from POAG patients followed for over 5 years.
• Global RNFL thickness measured annually using spectral-domain optical coherence tomography.
• Random forest (RF) model trained on systemic data to predict RNFL thinning rate, interpreted using SHAP values.

Main Results:

• The RF model achieved an R² of 0.88, accurately predicting RNFL thinning.
• Key systemic predictors identified: higher aspartate aminotransferase, lower blood glucose, lower systolic blood pressure, and higher HDL.
• Ophthalmic factors like higher baseline RNFL thickness and intraocular pressure were also significant predictors.

Conclusions:

• Baseline systemic features, including specific blood test results and blood pressure, are valuable predictors of POAG progression.
• Integrating systemic health data into predictive models can enhance glaucoma management strategies.
• This research highlights the systemic impact on POAG and offers insights for personalized patient care.