Relying on a combination of available evidence and expert opinion, a panel of ophthalmologists has reviewed nearly 150 statements pertaining to the evaluation and treatment of primary open-angle glaucoma and has reached consensus agreement or consensus disagreement with the majority of the statements. The results are intended to inform clinicians how glaucoma practitioners view this disease but are not guidelines or recommended practice patterns, according to several steering committee members of the project panel.
Relying on a combination of available evidence and expert opinion, a panel of ophthalmologists has reviewed nearly 150 statements pertaining to the evaluation and treatment of primary open-angle glaucoma (POAG) and has reached consensus agreement or consensus disagreement with the majority of such statements.
The panel's findings and methodology were published in two papers the March issue of the American Journal of Ophthalmology (2008;145:570-581). The results are intended to inform clinicians how glaucoma practitioners view this disease but are not guidelines or recommended practice patterns, according to several steering committee members of the project panel.
M. Roy Wilson, MD, chancellor, University of Colorado Denver, was co-chairman and one of the four members of the steering committee for the Modified RAND-like Methodology Group as well as lead author of the paper on methodology and design. The other co-chairman of the program was Kuldev Singh, MD, MPH, professor of ophthalmology and director of the glaucoma service at Stanford University, Stanford, CA.
Dr. Wilson concurred that the results of the panel's efforts synthesize much of the current thinking on glaucoma treatment.
"In terms of the amount of literature out there, we're bombarded on a daily basis with this article or that article and this claim or that claim. The average person probably receives five or six monthly journals, and you just can't distill all this information," Dr. Wilson said. "We think this helps distill for the practicing ophthalmologist the best of expert opinion-which is still very important and where most people get their information-with the available evidence."
"Evidence is either not available or incomplete to support much of what we do in the clinic. Even the highest-quality evidence may be limited by the artificial nature of clinical trials and other similar high-quality studies," Dr. Singh said. "There clearly is a place for clinical judgment or expert opinion to supplement the available evidence. The modified RAND-type methodology used by this panel is particularly useful in circumstances in which hard evidence is lacking or incomplete."
For this project, a voting panel of 10 ophthalmologists reviewed a series of statements on glaucoma treatment that was compiled by the non-voting steering committee, and it also reviewed studies and commentaries developed through an in-depth literature search by another non-voting glaucoma specialist. (Please see the journal articles for a complete list of the polling statements and results.)
The panelists voted privately through an online system, assigning a numeric value from 1 to 9 to each statement, with ratings of 1 and 9, respectively, representing strong disagreement or agreement with the statements. When the results were tallied, each statement was classified as having consensus agreement (the committee agreed that the statement was correct), consensus disagreement (the panel agreed that the statement was incorrect) or being indeterminate. Indeterminant statements were those with median scores of 4, 5, and 6 or ranges greater than 3 after the highest and lowest scores were discarded. This scoring system, not dissimilar to one used in Olympic competitions such as ice-skating, where subjective scoring can be problematic, minimized the effect of outliers.
Consensus agreement or disagreement was reached on 55.4% of the polling statements before the panel meeting and on 74.3% of the statements afterward. The panel failed to reach either consensus agreement or consensus disagreement on 38 statements.