Predicting visual outcomes in keratoprosthesis surgery with critical flicker fusion frequency, B-scan, visual electrophysiology and endoscopy.

Abstract

This study evaluates the predictive power of the critical flicker fusion frequency (CFF) test for visual outcomes in keratoprosthesis (KPro) candidates, comparing its accuracy with B-scan ultrasound, flash visual evoked potentials (fVEP) and endoscopy. The study included 42 patients (42 eyes) scheduled for KPro surgery with a median follow-up period of 6 months. The receiver operating characteristic curve identified the cut-off threshold for CFF in the model development study (17 eyes). All patients in the comparison study (25 eyes) underwent preoperative assessments including trichromatic CFF (red, green and yellow), B-scan ultrasound, fVEP and perioperative endoscopy. Results were classified as either favourable or unfavourable predictors of visual outcomes based on predefined criteria. Sensitivity and specificity of each assessment were calculated based on postoperative best-corrected visual acuity (BCVA)≥20/200. The Bland-Altman test assessed the consistency between CFF-predicted BCVA and actual BCVA. Among the trichromatic CFF tests, the yellow-CFF (yCFF) exhibited the highest area under the curve value of 0.97 and a cut-off threshold at 10 Hz for predicting postoperative BCVA≥20/200 (p<0.05). yCFF achieved 90% sensitivity and 80% specificity in predicting satisfactory postoperative outcomes. Endoscopy had 80% sensitivity and 80% specificity, B-scan showed 70% sensitivity and 60% specificity, and fVEP had 75% sensitivity and 40% specificity. yCFF showed a mean bias of 0.091 logarithm of the minimum angle of resolution (logMAR) in postoperative prediction. The CFF test provides robust visual function evaluation in KPro candidates. It demonstrates superior predictive accuracy for visual prognosis compared with routine ophthalmologic examinations, such as B-scan ultrasonography, fVEP and endoscopy.