Eight Methods to Detect Glaucoma Progression

The early detection of visual field progression is one of the most challenging tasks for the glaucoma specialist. Moreover, in the clinical setting, oftentimes different methods can present conflicting answers. When that occurs, is there one method you should lean on more than another?

Alessandro Rabiolo, MD, of the department of ophthalmology at the University Vita-Salute, IRCCS San Raffaele, Milan, Italy, said that, if pressed, he would defer to trend-based methodologies. His suggestion is based on the results of a study involving 253 individuals, which he presented at the American Academy of Ophthalmology’s 2018 annual meeting in Chicago.

Investigators assessed nearly 5,000 visual fields from 372 eyes in participants with primary open-angle glaucoma. Each patient had ≥6 visual field readings and were followed for at least 3 years. Researchers also ran computer-simulated visual fields. They tested these methods:

• Advanced Glaucoma Intervention Study (AGIS) scores
• Collaborative Initial Glaucoma Treatment Study (CITGS) scores
• Mean deviation (MD) rate
• Visual field index (VFI) rate
• Permutation of pointwise linear regression (PoPLR)
• Guided progression analysis (GPA)
• Expert clinician evaluation
• Glaucoma rate index (GRI), a method based on pointwise exponential regression

Among the results:

• Overall, the detection rate was highest with PoPLR (65%).
• This was followed by GRI (58%), GPA (41%) CIGTS (36%), expert evaluation (35%), MD rate (31%), AGIS (29%), and VFI rate (22).
• Specificity was highest for VFI rate and AGIS (each 93%).
• This was followed by MD rate (90%), CIGTS (71%), GRI (38%), and PoLPR (24%)
• GRI and PoPLR produced the shortest times to identifying progression.

Dr Rabiolo concluded that GRI was the best method when considering a blend of detection rate, specificity, and time to detection. He added that it was “always beneficial to consider 2 or 3 methods depending on your objectives.”

However, what about instances where different methodologies produce conflicting answers? In such cases, he would bow to trend-based methods such as MD and VFI rate. Why? Because “these give you the rate of change, so you are able to distinguish between people with a low rate of progression and those with a faster rate,” Dr Rabiolo said. Event-based methodologies usually provide just a yes or no answer to progression.   

Reference

Rabiolo A, Morales E, Capistrano E, et al. Comparison of eight methods to detect and measure glaucomatous perimetric progression. Talk presented at: AAO 2018 annual meeting; October, 26-30, 2018; Chicago.