30-Second Takeaway
- Major strabismus surgery complications are rare but vision-threatening and cluster in older patients and complex procedures.
- Failure risk after minimally invasive bleb surgery may rise in non-White patients and with lower mitomycin-C dosing.
- AI systems now show specialist-level performance for glaucoma and diabetic retinopathy diagnosis across ancestries and devices.
- Early structural damage in CSC during PDT delays highlights the need to avoid prolonged subretinal fluid exposure.
- Pathogen virulence in postprocedural endophthalmitis strongly predicts visual outcome and need for aggressive intervention.
Week ending December 6, 2025
Fresh Evidence with Near-Term Impact for Ophthalmic Practice
IRIS Registry Quantifies Major Complications After Strabismus Surgery
Among 89,339 patients undergoing strabismus surgery, major complications within 60 days occurred in 0.115%. Rates were 0.03% for corneal ulcers, 0.01% for endophthalmitis, and 0.04% each for orbital infection and lost or slipped muscle. Older age increased risk for all complications except orbital infection, and combined vertical–horizontal surgery plus more add-on codes were independent risk factors. Female sex and age over 65 independently increased endophthalmitis risk, and about half of infections presented within 5 days. Visual loss was common, averaging 0.58 logMAR in endophthalmitis and 0.11 logMAR across other complications, supporting an early postoperative visit within 5 days.
Systematic Review Flags Failure Risks After Minimally Invasive Bleb Surgery
This review pooled 19 cohorts (9,580 eyes) undergoing XEN Gel Stent or PreserFlo MicroShunt implantation. Non-White ethnicity, particularly Asian and non-Hispanic Black, was associated with higher failure risk, though certainty was low. Lower intraoperative mitomycin-C dosing or omission also correlated with higher failure, suggesting MMC titration is important. Associations between glaucoma subtype and failure were inconsistent, with some signal for higher risk in pseudoexfoliative, angle-closure, uveitic, and normal-tension glaucoma. Evidence quality and GRADE certainty were low to moderate, so surgeons should individualize MMC dosing while awaiting stronger prognostic data.
ChatGPT o1 Pro Approaches Specialist Performance for Glaucoma Detection
In 204 biobank participants, ChatGPT o1 Pro classified glaucoma using only OCT RNFL and visual field numeric outputs. Against consensus glaucoma specialists, sensitivity was 96.0% and specificity 83.7%, with an AUC of 0.899 and accuracy 85.2%. Negative predictive value was 99.3%, but positive predictive value was 45.3%, indicating strong rule-out but limited standalone rule-in utility. Diagnostic performance was similar across European and non-European ancestries and across high versus lower polygenic risk strata.
Network Meta-analysis Compares Unilateral Amblyopia Therapies
This network meta-analysis synthesized 22 randomized trials of unilateral amblyopia, including spectacles, patching regimens, atropine, and adjunctive approaches. Compared with spectacles alone, full-time patching plus video games produced significantly greater visual acuity gains (MD 0.27 logMAR). Combined atropine with patching also outperformed spectacles (MD 0.25 logMAR), indicating benefit from multimodal penalization. Other regimens, including shorter-duration patching or monotherapy atropine, did not show statistically significant superiority over spectacles.
References
Numbered in order of appearance. Click any reference to view details.
Additional Reads
Optional additional studies from this edition.