Purchase this article with an account.
Gabrielle Graves, Murtaza Adam, Kimberly E. Stepien, Dennis P. Han; Evaluation of the Color Difference Plot Scoring System Analysis of the 103 Hexagon Multifocal Electroretinogram in the Detection of Hydroxychloroquine Retinal Toxicity. Invest. Ophthalmol. Vis. Sci. 2012;53(14):5229.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To evaluate sensitivity, specificity, and reproducibility of color difference plot analysis (CDPA) of 103-Hexagon multifocal electroretinogram (mfERG) in detecting established hydroxychloroquine (HCQ) retinal toxicity.
23 patients taking HCQ were divided into nontoxic and toxic groups based on Humphrey 10-2 visual fields, fundus exam and/or OCT. The control patient group had 78 patients without retinal disease and not taking HCQ. Two masked examiners evaluated age-corrected mfERG in the central ring (Rc) (0 to 5.5 degrees from fixation) and paracentral ring (Rp) (5.5 to 11 degrees from fixation). An abnormal ring was defined as containing any hexagons with a difference of 2 or more standard deviations from normal (color blue or black). Test sensitivity and specificity were determined. The association between CDPA results and presence of toxicity was determined using Fisher exact test. The results were compared to ring ratios acquired by Adam et al for discordance using McNemar’s test. Intergrader agreement and reproducibility using CDPA were assessed using Cohen’s kappa.
Rc had 83% sensitivity and 93% specificity. Rp had 89% sensitivity and 82% specificity. The sensitivity and specificity of Rc or Rp combined was 89% and 80% respectively. The Fisher exact test for Rc was P = 0.0010 and for Rp was P = 0.0365. The intergrader agreement for Rc was "very good" (kappa = 0.918, SE = 0.046). The intergrader agreement for Rp was also "very good" (kappa = 0.897, SE = 0.045). Both graders showed very high reproducibility (kappa = 0.93 to 1.000, and SE = 0.000 to 0.038). There was complete agreement of detecting toxicity comparing Rp to the ring ratio R5/R4. In this population, ring ratio analysis of R5/R4 had 89% sensitivity and 95% specificity.
CDPA showed high intergrader agreement and reproducibility. In the detection of HCQ toxicity, Rp had higher sensitivity, but lower specificity compared to Rc. Combining Rc and Rp had equal sensitivity as Rp. In this same population CDPA was equally effective as R5/R4 at detecting HCQ toxicity, but with lower specificity.
This PDF is available to Subscribers Only