Premium
Early detection of changes in visual function in diabetes mellitus
Author(s) -
Ismail G. M.,
Whitaker D.
Publication year - 1997
Publication title -
ophthalmic and physiological optics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.147
H-Index - 66
eISSN - 1475-1313
pISSN - 0275-5408
DOI - 10.1046/j.1475-1313.1997.97807582.x
Subject(s) - diabetic retinopathy , medicine , retinopathy , diabetes mellitus , hue , visual acuity , ophthalmology , contrast (vision) , optometry , audiology , endocrinology , artificial intelligence , computer science
Purpose: Several measures of visual performance were used to investigate the possibility of a reliable, quick and low‐cost method of differentiating between diabetic patients with and without early retinopathy. Methods: Four groups of patients were recruited: a normal control group (30 patients); a non‐insulin dependent diabetic group without retinopathy (30 patients); a similar group with early retinopathy (30 patients); a similar group with more advanced retinopathy (14 patients). High‐ and low‐contrast LogMAR visual acuity, Pelli‐Robson contrast sensitivity and the Farnsworth‐Munsell 100‐Hue colour vision test were performed on all patients by an experimenter who was independent of the classification stage. Results: Comparison of mean test scores for each subject group reveals deficits in performance for all tests as a function of severity of diabetic retinal changes. However, none of the tests were successful in differentiating the subject groups with and without early retinopathy ( P >>0.05). Larger differences in performance were found when comparing the non‐diabetics against diabetics without retinopathy. Both the Pelli‐Robson chart and the 100‐Hue test result in mean scores which were significantly different in these two groups ( P < 0.01 and P < 0.05 respectively). QROC analysis of the results confirmed this impression, with kappa coefficients for efficiency of discriminability being generally higher for comparison between non‐diabetics and those without retinopathy (0.17 < k < 0.37) than between early diabetic groups (0.13 < k < 0.26). Discrimination between patients with pre‐clinical diabetics and those with moderate retinopathy was more efficient (0.21 < k < 0.62). Conclusion: Tests of visual function appear to be of limited value in screening for the presence of early retinopathy. Rather, a mild decline in visual function appears to already be present in those diabetics without retinopathy.