Characterization of Temporal Delay and Modeling of Intraocular Pressure and Visual Field Progression in the African...

Despite its utility, mathematical modeling of an individualized intraocular pressure (IOP) goal has been hindered by the variable temporal relationships between IOP and visual field (VF) rate of progression (ROP). In this study, we fit a regression model after accounting for the temporal delays between IOP and ROP to help predict individual IOP goals.

This is a retrospective study analyzing the IOP-ROP temporal relationship and identifying predictive factors of IOP of stability (IOS).

Four hundred eight eyes from the African Descent and Glaucoma Evaluation Study with more than 14 VFs were included.

For each eye, a ROP time series was produced and correlated to the IOP time series with incremental temporal shifts to estimate the delay between the 2 series. Average IOP during VF progression (worse than -0.5 dB/year) and stability (better than -0.5 dB/year) were calculated accounting for the delay. Contributing factors for IOS were analyzed using multivariable analysis, and significant factors were used to fit a regression model to help predict an individualized IOP goal.

Temporal delay between IOP and ROP changes, and magnitude of error between the predicted IOS and the actual IOS.

Delay between IOP and VF changes was found to have a median of 7 months and mode of 4 months. Multivariable analysis found that thinner central corneal thickness (CCT) (P < 0.001) and greater mean deviation (MD) (P < 0.001) were associated with lower required IOP to maintain VF stability. Using CCT and MD, fitted regression models were able to predict individualized IOP that maintains VF stability within ±2 to 3 mmHg.

Rate of progression changes trail IOP changes by 4-7 months, an important consideration when obtaining additional VFs to capture or rule-out progression. Accounting for this delay, we modeled individualized IOS to within 2-3 mmHg using only 2 clinical measures, CCT and MD. This simple IOP goal calculation will enable clinicians to make early treatment decisions prior to the delayed VF-based confirmation.

Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.