PURPOSE: To determine the rate of agreement among five retina specialists in classifying various angiographic features of subfoveal choroidal neovascularization (CNV) secondary to age-related macular degeneration (AMD), as evaluated on printed digital fluorescein angiogram (FA) frames, as well as determination of eligibility for photodynamic treatment (PDT) according to established guidelines. METHODS: Ninety-two digital FAs demonstrating subfoveal CNV secondary to AMD were evaluated independently by five retina specialists. The pattern of classic component could be classified as no classic component, minimally classic, predominantly classic, or classic only. Each grader was asked to determine eligibility of each case to PDT according to established treatment guidelines, national health insurance guidelines, and one's own personal judgment. RESULTS: The kappa coefficient of concordance calculated for all five observers regarding CNV localization was 0.285, indicating fair overall agreement, and was 0.295, indicating fair agreement, regarding classification of leakage pattern. The kappa coefficient of agreement calculated for all five graders regarding eligibility for treatment according to established international guidelines, national health insurance, and each grader's own personal judgment was 0.163, 0.33, and 0.164, respectively, indicating slight to fair overall agreement. CONCLUSION: Considerable variability may exist among retina specialists interpreting FAs and should be considered in the assessment of treatment guidelines.
PURPOSE: To determine the rate of agreement among five retina specialists in classifying various angiographic features of subfoveal choroidal neovascularization (CNV) secondary to age-related macular degeneration (AMD), as evaluated on printed digital fluorescein angiogram (FA) frames, as well as determination of eligibility for photodynamic treatment (PDT) according to established guidelines. METHODS: Ninety-two digital FAs demonstrating subfoveal CNV secondary to AMD were evaluated independently by five retina specialists. The pattern of classic component could be classified as no classic component, minimally classic, predominantly classic, or classic only. Each grader was asked to determine eligibility of each case to PDT according to established treatment guidelines, national health insurance guidelines, and one's own personal judgment. RESULTS: The kappa coefficient of concordance calculated for all five observers regarding CNV localization was 0.285, indicating fair overall agreement, and was 0.295, indicating fair agreement, regarding classification of leakage pattern. The kappa coefficient of agreement calculated for all five graders regarding eligibility for treatment according to established international guidelines, national health insurance, and each grader's own personal judgment was 0.163, 0.33, and 0.164, respectively, indicating slight to fair overall agreement. CONCLUSION: Considerable variability may exist among retina specialists interpreting FAs and should be considered in the assessment of treatment guidelines.
Authors: Michael F Chiang; Sarah Read-Brown; Daniel C Tu; Dongseok Choi; David S Sanders; Thomas S Hwang; Steven Bailey; Daniel J Karr; Elizabeth Cottle; John C Morrison; David J Wilson; Thomas R Yackel Journal: Trans Am Ophthalmol Soc Date: 2013-09
Authors: Sobha Sivaprasad; Katie Banister; Augusto Azuro-Blanco; Beatriz Goulao; Jonathan A Cook; Ruth Hogg; Graham Scotland; Heinrich Heimann; Andrew Lotery; Faruque Ghanchi; Richard Gale; Geeta Menon; Louise Downey; Nicola Hopkins; Peter Scanlon; Ben Burton; Craig Ramsay; Usha Chakravarthy Journal: Ophthalmology Date: 2021-07-28 Impact factor: 12.079