Purchase this article with an account.
Julia Nemiroff, Carlos G. De Moraes, Christopher C. Teng, Sung Chul Park, Celso Tello, Robert Ritch, Jeffrey M. Liebmann; Inter-observer Agreement in Detecting Glaucomatous Progression Using Stereophotography. Invest. Ophthalmol. Vis. Sci. 2011;52(14):4143.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To investigate agreement among glaucoma specialists in identifying structural progression in glaucoma patients using stereophotography.
Glaucoma patients with repeatable VF loss and good quality stereophotographs in either eye evaluated between January 1999 and December 2009 were included. Optic disc photos were reviewed by two glaucoma specialists masked as to their temporal sequence and clinical/VF information. Photos were presented in a computer screen with alternating images using a stereoviewer. Photos were graded whether progression occurred, form of progression (rim or RNFL loss, PPA enlargement, or DH), and location (superior, inferior, or diffuse). Disagreements were adjudicated by a third grader. False-positive rates were determined in cases of progression agreement but with mistaken interpretation of the temporal sequence.
389 eyes (389 patients; mean age 64.9±13.0 yrs; mean baseline MD, -7.1±5.1) were investigated. There was a moderate agreement between the two graders regarding structural progression (72%, kappa=0.42). Following adjudication, 115 eyes (29.5%) were depicted as presenting structural progression. 33 eyes (28.6%) were false-positive results; therefore they were excluded from further analyses. Regarding the type of progression, rim change and DH detection yielded the best inter-observer agreement (75.6% each), followed by RNFL change (61%), and PPA enlargement (59%). Graders agreed regarding the location of most prominent structural change in 53.6% of cases. Among these eyes, the most common location of progression was diffuse (66%), followed by inferior (18%) and superior (16%). There was no significant difference in the time between the first and last evaluated photograph between cases of agreement and disagreement (6.6±1.6 vs. 6.4±1.6 yr, p=0.60) nor in their global velocities of VF progression (-0.43±0.6 vs. -0.42±0.6 dB/yr, p=0.26). However, cases of disagreement tended to present worse baseline MD than cases of agreement (-7.3±4.8 vs. -6.3±4.8 dB, p=0.05, Mann-Whitney test).
We found a moderate agreement among observers to detect structural progression using stereophotography. Disagreement between observers is more likely to occur in eyes with moderate to severe VF loss.
This PDF is available to Subscribers Only