Purchase this article with an account.
Mariel Lavieri, Jonathan Helm, Greggory Schell, Mark Van Oyen, David Musch, Joshua Stein; Personalizing the Frequency and Timing of Testing to Check for Glaucoma Progression: a Novel Approach. Invest. Ophthalmol. Vis. Sci. 2013;54(15):3956.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To determine whether it is possible to significantly improve detection of disease progression in patients with open angle glaucoma (OAG) by using personalized, dynamically-adjusted schedules of visual field (VF) testing and intraocular pressure (IOP) measurements.
A Kalman filter was used to model the disease dynamics of a large group of patients with mild to advanced OAG based on VF and IOP readings obtained sequentially over time. As additional VF and IOP tests are performed on each patient, the filter updates our knowledge about each patient’s disease dynamics. Parameterization and validation of the models was performed using >5 years of longitudinal data of IOP measurements and VF testing from patients who were enrolled in the Collaborative Initial Glaucoma Treatment Study (CIGTS) and the Advanced Glaucoma Intervention Study (AGIS). The model then forecasts each patient’s disease dynamics into the future while incorporating the uncertainty associated with those forecasts. Logistic regression was used to model the relationship between the current and future disease dynamics and OAG progression. We developed an algorithm which combines the Kalman filter learning capabilities and the logistic regression predictive power to determine personalized schedules of VF and IOP testing for each patient. Our algorithm was compared against fixed interval schedules for obtaining VFs and IOP measurements from the trials.
A total of 571 participants (571 eyes) with OAG were evaluated. Compared to yearly intervals of checking VFs and IOPs, our approach using the Kalman filter achieved a 27% increase of efficiency in detecting OAG progression (p<0.0001) and detected OAG progression 63% sooner (i.e. reduced diagnostic delay) (p<0.0001) using the same number of tests per patient.
Dynamic and personalized schedules for obtaining IOP and VF measures using a Kalman filter approach can improve the likelihood of detecting OAG progression and identifying OAG progression sooner than fixed interval examination schedules.
This PDF is available to Subscribers Only