Agreement between ophthalmologists and optometrists in optic disc assessment: training implications for glaucoma co-management

Graefes Arch Clin Exp Ophthalmol. 2001 Jun;239(5):342-50. doi: 10.1007/s004170100272.


Background: Many studies have estimated observer variability for optic disc assessment among experts, but there are few data on agreement between optometrists and ophthalmologists. The aim of this study is to report inter- and intraobserver agreement among optometrists and ophthalmologists in optic disc assessment and discuss the implications for glaucoma co-management.

Methods: Twelve observers (six optometrists and six ophthalmologists) graded 48 stereo-pairs of optic disc photographs from 48 patients on two separate occasions. Observers graded the vertical C/D ratio, the narrowest rim width and the presence/absence of disc haemorrhage. Agreement was assessed by calculating the standard deviation (SD) of differences and the kappa statistic, within and between observers. Systematic differences in grading between and within individuals were described by calculating mean differences.

Results: Intra-observer agreement is "substantial" and significantly better than inter-observer agreement for all disc features (P<0.0001). Intra-observer agreement is comparable for optometrists and ophthalmologists with regard to C/D ratio estimates. Although some optometrists show close agreement with ophthalmologists, overall agreement between optometrists and ophthalmologists is significantly worse than agreement among ophthalmologists alone for vertical C/D ratio (P=0.002) and disc haemorrhage (P=0.02). There are fewer inter-observer mean differences that differ significantly from zero among ophthalmologists (7%) than among optometrists (67%; chi2=11.63, P<0.001) or between optometrists and ophthalmologists (69%; chi2=16.7, P<0.001). There is evidence of systematic under-reading of C/D ratios by less experienced optometrists.

Conclusion: Although individual optometrists can show good agreement with ophthalmologists on disc grading, the variation in individual performance indicates that training and accreditation in disc assessment is an essential prerequisite for participation by optometrists in glaucoma co-management.

Publication types

  • Comparative Study

MeSH terms

  • Clinical Competence / standards*
  • Glaucoma / diagnosis*
  • Humans
  • Observer Variation
  • Ophthalmology / education*
  • Optic Disk / pathology*
  • Optometry / education*
  • Patient Care Team
  • Reproducibility of Results