Y-split versus standard lateral rectus recession for sensory exotropia: a propensity-score adjusted cohort study.

PURPOSE: The Y-split lateral rectus recession has biomechanical advantages in incomitant strabismus, but its role as a primary weakening procedure for comitant large-angle exotropia is unclear. We compared surgical success and mechanical efficiency between Y-split and standard recession in predominantly sensory exotropia. METHODS: This retrospective study included 86 patients (41 standard, 45 Y-split; recession 7-9 mm). Success was a 6-month distance deviation < 10 PD. Non-random allocation was addressed with inverse-probability weighting (IPTW), propensity-score matching, overlap weighting, entropy balancing, and recession adjustment. Eyes were classified as sensory exotropia or divergence excess by acuity criteria. An E-value assessed unmeasured confounding. Reported per STROBE. RESULTS: The Y-split group had larger deviations (median 40 vs. 30 PD, p < 0.001) and poorer acuity (logMAR 1.0 vs. 0.1, p = 0.001). Unadjusted success was 48.9% (Y-split) versus 36.6% (standard, p = 0.35); no consecutive esotropia occurred. After IPTW, success odds were higher with Y-split (OR 4.75, 95% CI 1.86-12.12, p = 0.001). Corrective efficiency was 3.4-3.9 versus 2.4-2.7 PD/mm. The advantage persisted with overlap weighting (OR 6.49), recession adjustment (OR 5.66), across deviation strata (Mantel-Haenszel OR 11.0), and within the sensory exotropia subgroup (OR 4.54); all remained significant after false-discovery-rate correction. Complications were not systematically recorded. CONCLUSION: Y-split lateral rectus recession provided 30-40% more correction per millimetre and four- to five-fold higher adjusted odds of short-term success than standard recession, even in eyes with worse prognosis. It is a candidate procedure for large-angle sensory exotropia on preliminary evidence, requiring prospective confirmation with safety monitoring and at least 24 months of follow-up.