Predicting Future Self-Reported Motor Vehicle Collisions in Subjects with Primary Open-Angle Glaucoma Using the...

Area of Science:

• Ophthalmology
• Neurology
• Public Health

Background:

• Primary open-angle glaucoma (POAG) is a leading cause of irreversible blindness.
• Visual impairment in POAG patients can significantly impact driving safety.
• Predicting motor vehicle collisions (MVCs) in this population is crucial for public safety.

Purpose of the Study:

• To predict the likelihood of future MVCs in POAG patients.
• To evaluate the utility of visual function, driving attitudes, and past MVC history in predicting future MVCs.
• To apply the penalized support vector machine (pSVM) model for prediction.

Main Methods:

• 185 POAG patients were analyzed prospectively.
• Self-reported MVCs were collected for 3 years via surveys.
• Integrated visual fields (IVFs) were calculated, and divided into six regions.
• The pSVM model was used to predict future MVCs, incorporating variables like visual acuity, visual field data, driving attitudes, and past MVC history.

Main Results:

• The pSVM model incorporating driving attitudes (pred_{penSVM_all}) showed a significant relationship with future MVCs (OR = 1.21, P = 0.0015).
• The model without driving attitudes (pred_{penSVM_basic}) approached significance (OR = 1.15, P = 0.064).

Conclusions:

• The pSVM model is useful for predicting future MVCs in POAG patients.
• Integrating visual function, driving attitudes, and past MVC history enhances predictive accuracy.
• Careful consideration of these factors is essential for risk assessment in POAG patients.