Validity and Reliability of Semiautomatic Ocular Cycloposition Measurement With Spectralis Optical Coherence Tomography

Abstract

Disc-fovea angle determined by fundus photography (P-DFA) is considered the gold standard for cycloposition assessment. Fovea-to-disc alignment (FoDi) software of the spectral-domain optical coherence tomography (SD-OCT) (Spectralis) also measures the DFA (O-DFA) based on subject fixation and offers important clinical advantages. This study aimed to analyze the validity and reliability of measuring cycloposition using OCT and to determine its performance in eyes with poor foveal definition. Validity and reliability analysis. In 60 eyes with normal foveal definition and 32 eyes with poorly defined fovea, ocular cycloposition was assessed by 2 observers using 5 fundus photographs and 5 FoDi analyses each. Patients were repositioned after every capture. Cycloposition assessed by O-DFA was 7.6 ± 3.5-degrees, and P-DFA was 7.9 ± 3.8-degrees. The concordance between methods was good (intraclass correlation coefficient [ICC], 0.71), with absolute differences ranging from zero to 4-degrees in 85% of the subjects. The precision was 1.4-degrees for O-DFA and 3.0-degrees for P-DFA. Repeatability and reproducibility were excellent in both techniques. In the group of patients with poor foveal definition, the precision of P-DFA changed from 3.0-degrees to 4.8-degrees, whereas the O-DFA remained stable. OCT had a good agreement with the fundus photography method. O-DFA showed better precision than P-DFA. O-DFA repeatability and reproducibility were excellent and unconditioned by foveal status.