Clinical quality assessment using computer monitor photoimages of optic nerve head cupping.

Abstract

Photo documentation (color slides) of abnormal optic nerve heads enable the clinician to accurately access or estimate the percentage of the optic nerve cupping to quantify the progression of glaucoma. The purpose of this study is to compare cupping estimations using digital images on a computer monitor to those using standard color slides. Ten sets of stereo optic nerve head photos from six glaucoma suspects were used in this study. A set is two 35-mm pictures of one optic nerve head. The film was developed as standard color slides using Seattle Filmworks slide developing. The computer monitor photoimages used Seattle Photoworks software for photos on disk. Thirteen experienced examiners estimated cupping on the 10 sets of color slides and on the 10 sets of digital images. Standard criterion was used to estimate the C/D percentage. We used a three-way mixed model ANOVA experimental design to analyze the data. Examiner estimations of C/D differed slightly between color slides and computer monitor photoimages (mean C/D of 61.1% for slides to 63.7% for monitor). This difference was statistically significant (p = 0.033). The differences authors found were not clinically significant (10%) in 7 out of the 10 optic nerve head photos using the accepted standard of interexaminer reliability. The variations of the examiner estimations as well as interexaminer differences will be discussed. The use of a computer image photodisplay of an optic nerve head has a potential as an alternative to traditional slide photodocumentation. This study identifies problems which require more work to make these computer images the equivalent of color slides.