Design and validation of a method to determine the position of the fovea by using the nerve-head to fovea distance of the fellow eye.

Purpose: To measure the nerve-head to fovea distance (NFD) on fundus photographs in fellow eyes, and to compare the NFD between fellow eyes.

Methods: Diabetic patients without retinopathy, (n = 183) who were screened by fundus photography at the University Medical Center Groningen, the Netherlands from January 1(st) 2005 until January 1(st) 2006 were included. The NFD was measured in left and right eyes both from the center and from the rim of the nerve-head. To determine inter- and intra-observer agreement, repeated measurements by one observer (n = 3) were performed on all photographs and by two observers on 60 photographs (30 paired eyes). The effect of age, gender, and refractive error on NFD was analysed.

Results: The correlation of NFDs between the left and the right eye was 0.958 when measured from the center of the nerve head (mean difference 0.0078 mm. ±SD 0.079 (95% limits of agreement -0.147-0.163)) and 0.963 when measured from the rim (mean difference 0.0056±SD 0.073 (95% limits of agreement -0.137-0.149)). Using the NFD between fellow eyes interchangeably, resulted in a standard error of 0.153 mm. Intra- and inter-observer variability was small. We found a significant effect of age (center of the nerve-head (P = 0.006) and rim of the nerve head (P = 0.003)) and refractive error (center of nerve-head (P<0.001) and rim of nerve head (P<0.001)) on NFD.

Conclusions: The NFD in one eye provides a confident, reproducible, and valid method to address the position of the fovea in the fellow eye. We recommend using the NFD measured from the center of the nerve-head since the standard error by this method was smallest. Age and refractive error have an effect on NFD.