The Influence of Clinical Data on Cup/Disc Ratio and Disc Damage Likelihood Scale

Abstract

Purpose: : To assess the effect of ancillary data on disc evaluations using the vertical cup/disc ratio (CDR) and the Disc Damage Likelihood Scale (DDLS) systems.

Methods: : A set of stereo disc photographs from 36 glaucoma patients from the Glaucoma Service Diagnostic Laboratory at Wills Eye Institute was assembled. All cases had glaucomatous nerve damage with or without visual field loss. Three glaucoma fellows evaluated the set of 36 stereo disc photographs. All three reviewers interpreted the entire set of photographs 11 times using two different grading systems - CDR and the DDLS. The order of the photographs was shuffled between each reading and the reviewers were masked to the results of prior readings. For 3 of the readings, the photographs were examined without any accompanying material or information. For the other 8 readings, various amounts of ancillary clinical information, either correct or misleading, was shown to the masked readers at the time they were evaluating the 36 photographs. The baseline level of agreement for grading the photos was established and the impact of correct or erroneous clinical history on the level of agreement was measured.

Results: : The baseline level of agreement between the observers using the CDR was from 91% to 94% and for the DDLS was from 82% to 86%. When the correct clinical information was provided, the level of agreement increased from baseline by 1% to 4% using the CDR, and by 2% to 7% using the DDLS. Conversely, when incorrect clinical data was provided, the level of agreement decreased by 1% to 10% using the CDR and 12% to 22% using the DDLS. A correct clinical history significantly impacted the odds of agreement with either scoring system (CDR: OR=0.30, P=0.019; DDLS: OR=0.31, P=0.001). Visual field and intraocular pressure (IOP) were the factors most likely to impact disc interpretation when using the both systems.

Conclusions: : For both grading systems, ancillary clinical data has an impact on optic nerve interpretation. The visual field and IOP seem to influence this evaluation the most. The DDLS appears more likely to be influenced by clinical information than CDR.

Keywords: clinical research methodology • optic disc