Philip Yu, Kimia Mofid, Faizan Naveed, Lia Huo, Brendan K Tao, Austin Pereira, Tina Tang, Wai-Ching Lam, Efrem D Mandelcorn, David Wong, Peng Yan
Low-to-moderate-certainty evidence supports an incidence of unexplained vision loss following SO removal of approximately 1 in 25 eyes among reported cohorts, with substantial heterogeneity. Because eligible studies were selected for reporting this complication, this figure should be interpreted as an upper-bound rather than a generalisable population-level risk. Clinicians should consider minimising SO tamponade duration when feasible and obtain imaging before and after removal to detect early structural changes.
BACKGROUND/OBJECTIVES: Sudden, unexplained visual acuity loss has been reported following silicone oil (SO) removal. This systematic review and meta-analysis aimed to estimate the pooled incidence of unexplained vision loss following SO removal (UVLSOR) and identify associated risk factors.
SUBJECTS/METHODS: We searched Ovid MEDLINE, Embase, and Cochrane Library from inception to November 2025 for studies reporting unexplained vision loss after SO removal following vitreoretinal surgery. Paired independent reviewers conducted screening, data extraction, and risk of bias assessment using the ROBINS-I tool. The pooled incidence was calculated using a random-effects meta-analysis.
RESULTS: Eight retrospective cohort studies (n = 1531 eyes; 54 events) were included. The pooled incidence of UVLSOR was 4.38% (95% CI: 2.42-7.82%; I² = 72.9%). Heterogeneity was substantial and largely reflected differences in underlying indication, tamponade characteristics, and case definitions; because only studies that specifically reported this complication were eligible, the pooled value likely overestimates the true population-level incidence. Vision loss typically occurred within 1-3 months post-removal and was defined as a loss of ≥2 Snellen lines without identifiable cause. Associated factors included longer SO tamponade duration, peri-tamponade ocular hypertension, and pre-removal indicators of inner retinal dysfunction on optical coherence tomography or multifocal electroretinography. Risk of bias was predominantly low-to-moderate, and certainty of evidence ranged from low to moderate.
CONCLUSIONS: Low-to-moderate-certainty evidence supports an incidence of unexplained vision loss following SO removal of approximately 1 in 25 eyes among reported cohorts, with substantial heterogeneity. Because eligible studies were selected for reporting this complication, this figure should be interpreted as an upper-bound rather than a generalisable population-level risk. Clinicians should consider minimising SO tamponade duration when feasible and obtain imaging before and after removal to detect early structural changes.