Disc-damage Likelihood Scale (DDLS) as a Clinical Indicator of the Presence of a Relative Afferent Pupillary Defect (RAPD)

• Published in: Journal of glaucoma Vol. 25; no. 10; p. e910
• Main Authors: Zhang, Alice Y, Lu, Lan, Ali, Mohsin, Rutnin, Nont, Faria, Bruno M, Guzel, Huseyin, Liang, Liang, Martinez, Patricia, Tawfik, Mike, Spaeth, George L
• Format: Journal Article
• Language: English
• Published: United States 01.10.2016
• Subjects: Adolescent; Adult; Aged; Aged, 80 and over; Female; Humans; Intraocular Pressure - physiology; Male; Middle Aged; Optic Disk - pathology; Optic Nerve Diseases - diagnosis; Probability; Pupil Disorders - diagnosis; Retinal Ganglion Cells - pathology; Retrospective Studies; Sensitivity and Specificity; Tomography, Optical Coherence; Tonometry, Ocular; Visual Acuity - physiology; Visual Fields - physiology; Young Adult
• ISSN: 1536-481X
• DOI: 10.1097/IJG.0000000000000425

To evaluate clinical parameters and their relationship to the presence of a relative afferent pupillary defect (RAPD). Retrospective chart review of 672 consecutive patients who presented to the Glaucoma Service at Wills Eye Hospital from January 1 through May 29, 2012. Swinging flashlight method (SFM) was used to test for RAPDs. Records of visual acuity, intraocular pressure (IOP), disc-damage likelihood scale (DDLS), cup/disc (C/D) ratio, visual field mean deviation (MD), optical coherence tomography (OCT), and Heidelberg retinal tomography (HRT) asymmetries were examined. We measured the prevalence of RAPDs as clinical asymmetries increase, calculated cut-off points for clinical asymmetries that optimized sensitivity and specificity in detecting RAPDs, and determined values of clinical asymmetries above which a RAPD always exists. Upon exclusion, we studied 409 patients, 175 (42.8%) with RAPDs and 234 (57.2%) without. Age, visual acuity, IOP, DDLS, C/D ratio, MD, retina nerve fiber layer thickness by OCT, HRT C/D, and HRT rim area asymmetries all correlated with RAPD status (OCT and HRT parameters did not include enough patients for multivariable analysis or cut-off determination). Multivariable analysis indicated that IOP, DDLS, and MD asymmetries were significantly correlated with RAPD status (P-value<0.05). Although the optimal cut-off values for the variables retained in the final multivariable model had low sensitivity and moderate specificity, DDLS had the highest specificity of 0.86. IOP, DDLS, and MD asymmetries had the best correlation with RAPD status, and increased asymmetries in these parameters were associated with higher likelihood of RAPDs. DDLS score had the highest specificity in predicting a RAPD, and DDLS asymmetry scores ≥6 identified all cases of RAPDs.