Sensitivity of ultra-widefield fundus photography for retinal break localisation in primary rhegmatogenous retinal detachment

To determine the sensitivity of ultra-widefield imaging (UWFI) for detecting retinal breaks in primary rhegmatogenous retinal detachment (RRD) and to evaluate its utility for preoperative localisation.

We included adults with primary RRD who underwent PPV and had preoperative UWFI. Intraoperative scleral indentation served as reference standard. Breaks were classified by morphology and by a fovea-centred, right-eye standardised clock-hour system. The primary outcome was causative-break detection on UWFI and quadrant-level sensitivities were secondary.

Among 363 eyes with primary RRD, 9 were excluded for vitreous haemorrhage, leaving 354 eyes for analysis. The causative break was detected on UWFI in 208 eyes (58.8%). All breaks were identified on UWFI in 148 eyes (41.8%). Compared with the not-detected group, the detected group was younger (57.8 ± 9.5 vs 61.3 ± 11.6, P = 0.003), had a higher phakic rate (84.6% vs 78.0%, P = 0.041), and showed a smaller detachment extent (1.88 ± 0.72 vs 2.08 ± 0.87 quadrants, P = 0.025). Horseshoe tears were more common in the detected group (80.3% vs 57.5%, P < 0.001), while round holes were more frequent in the not-detected group (42.5% vs 16.8%, P < 0.001). Overall sensitivity for all breaks was 49.1% (308/627). Detection was highest in the temporal quadrant (69.2%) and lower in the nasal (45.9%), superior (45.2%), and inferior (33.0%) quadrants. Clock-hour analysis peaked at 9 to 10 o'clock and was lowest at 5 to 6 o'clock, indicating better temporal and poorer inferior detection.

UWFI alone has limited sensitivity for detecting retinal breaks in patients undergoing retinal detachment repair. UWFI may complement but not replace scleral indentation in perioperative evaluation.