Digital Optic Nerve Photography in Assessing CuptoDisc Ratio: The Experience in an Academic Institut

1
Digital Optic Nerve Photography in Assessing
Cup-to-Disc Ratio The Experience in an Academic
Institution
Naomi Goldberg, Lynda Z. Kleiman, Eric J.
Wolf Edward S. Harkness Eye Institute Columbia
University College of Physicians and Surgeons New
York, NY
2
Purposes

• To compare the level of agreement among
  ophthalmic specialists in an academic institution
  in estimating cup-to-disc ratios from digital
  disc photographs.
• To assess whether ophthalmology residents at
  different levels of training improve in
  interpreting cup-to-disc ratios from digital disc
  photographs.
• 3. To assess whether cup-to-disc ratios based on
  disc photographs are sufficient in communicating
  clinical data between practitioners.

3
Materials and Methods

• Digital optic nerve head images were obtained of
  five eyes from glaucoma or glaucoma-suspect
  patients and five eyes from normal patients.
  First, second, and third year ophthalmology
  residents, as well as attending ophthalmologists
  from the Cornea Service, the Retina Service, the
  Glaucoma Service and the Pediatrics Service were
  individually asked to determine vertical
  cup-to-disc ratios from the ten images.

4
Figure 1a
Figure 1b
5
Figure 2a
Figure 2b
6
Figure 3a
Figure 3b
7
Figure 4
8
Conclusions

• Among individual examiners, the maximal
  difference between estimates is greater for
  glaucoma eyes (0.25-0.35) than for control eyes
  (0.2-0.3).
• Individual examiners more strongly agreed in
  their assessments of control eyes rather than
  glaucoma eyes, and agreement was strongest
  regarding Eye H (standard deviation 0.06). Thus,
  there is strongest consensus in assessing
  non-glaucomatous nerves, especially those with
  smallest cup-to-disc ratios.
• 3. Individual examiners disagreed most in their
  assessments of the cup-to-disc ratio in Eye B
  (standard deviation 0.09). Among glaucoma eyes,
  however, examiners agreed most on Eye D (standard
  deviation 0.07). Thus, glaucomatous nerves were
  subject to the most disagreement, although there
  is a suggestion that there is less discrepancy
  with more highly cupped nerves.

9
Conclusions - continued

• 4. There is more overall agreement between
  residents at different stages of training than
  there is between specialists in determining
  cup-to-disc ratios.
• The discrepancies between sub-specialists
  estimates are more pronounced in glaucomatous
  eyes.
• There is no discernible trend in estimation of
  cup-to-disc ratio that is unique to a particular
  subspecialty i.e. we cannot conclude that one
  group of specialists consistently over- or
  under-estimates cupping as compared to other
  specialists.
• 5. There was no trend toward graded improvement
  in residents assessments of cup-to-disc ratios.
• 6. In test eyes, residents tend to overestimate
  the degree of cupping relative to attending
  physicians.
• 7. The data here suggest that cup-to-disc ratio
  based on optic nerve photography may not be
  sufficient in communicating clinical information
  between practitioners. Further studies should be
  done to further clarify this issue.