BACKGROUND AND OBJECTIVE: Retinal nonperfusion is closely associated with vision-threatening complications such as neovascularization and macular edema. The purpose of this study is to investigate the reliability of a calculated ischemic index (ISI) by means of intergrader and intragrader agreement on ultrawide-field fluorescein angiography (UWFFA) in common retinal vascular diseases. PATIENTS AND METHODS: Eight trained graders evaluated 15 UWFFA images provided digitally and re-graded on a different day. They included five eyes with diabetic retinopathy (DR), five with branch retinal vein occlusion (BRVO), and five with central retinal vein occlusion (CRVO). To assess intergrader and intragrader agreement and variability among different diseases, the replicate inter- and intragrader standard deviations (SDs) and coefficients of variation (CVs) were calculated. RESULTS: Mean ISI was 46% for images of DR, 26% for images of BRVO, and 61.3% for images of CRVO. Combined intragrader and intergrader replicate SDs were 17.8% for DR, 3.8% for BRVO and 13.0% for CRVO. Combined intragrader and intergrader replicate coefficients of variation were 38.6% (percent of mean ISI) for DR, 14.7% for BRVO, and 21.2% for CRVO. CONCLUSION: Intergrader and intragrader variability was high when assessing DR. This may be due to the chronic nature of DR progression, which can lead to patchy areas of ischemia. Intergrader and intragrader variability was better for CRVO and best for BRVO. This may be due to the acute or subacute nature of retinal vein occlusions. Copyright 2015, SLACK Incorporated.
BACKGROUND AND OBJECTIVE: Retinal nonperfusion is closely associated with vision-threatening complications such as neovascularization and macular edema. The purpose of this study is to investigate the reliability of a calculated ischemic index (ISI) by means of intergrader and intragrader agreement on ultrawide-field fluorescein angiography (UWFFA) in common retinal vascular diseases. PATIENTS AND METHODS: Eight trained graders evaluated 15 UWFFA images provided digitally and re-graded on a different day. They included five eyes with diabetic retinopathy (DR), five with branch retinal vein occlusion (BRVO), and five with central retinal vein occlusion (CRVO). To assess intergrader and intragrader agreement and variability among different diseases, the replicate inter- and intragrader standard deviations (SDs) and coefficients of variation (CVs) were calculated. RESULTS: Mean ISI was 46% for images of DR, 26% for images of BRVO, and 61.3% for images of CRVO. Combined intragrader and intergrader replicate SDs were 17.8% for DR, 3.8% for BRVO and 13.0% for CRVO. Combined intragrader and intergrader replicate coefficients of variation were 38.6% (percent of mean ISI) for DR, 14.7% for BRVO, and 21.2% for CRVO. CONCLUSION: Intergrader and intragrader variability was high when assessing DR. This may be due to the chronic nature of DR progression, which can lead to patchy areas of ischemia. Intergrader and intragrader variability was better for CRVO and best for BRVO. This may be due to the acute or subacute nature of retinal vein occlusions. Copyright 2015, SLACK Incorporated.