Simple Scoring Model for Evaluation of Traumatic Optic Neuropathy in Unconscious Patients with Periorbital Facial Bone Fracture

Background: Traumatic optic neuropathy (TON) is a rare but severe optic nerve injury resulting from periorbital trauma. Diagnosing TON is particularly challenging in unconscious patients, making early detection crucial to prevent irreversible vision loss. Methods: We conducted a retrospective review of patient data from 2008 to 2020, focusing on individuals with periorbital facial bone fractures. Independent risk factors for TON were identified through multivariable logistic regression analysis on data from May 2008 to April 2019. A prediction scoring model was subsequently developed and validated with data from May 2019 to April 2020. Results: A total of 2008 eyes with periorbital facial bone fractures were analyzed (development cohort: 1639 eyes; validation cohort: 369 eyes). Among these, 97 eyes (5.92%) in the development cohort were diagnosed with TON. The scoring model included: optic canal fracture (3 points), medial orbital wall fracture (2 points), orbital roof fracture (1 point), intracranial hemorrhage (1 point), and retrobulbar hematoma (2 points). The model demonstrated good discrimination (area under the receiver operating characteristic curve [AUC]: 0.84) and calibration (Hosmer-Lemeshow test, P = 0.70). Validation confirmed strong performance with an AUC of 0.87. Conclusion: The developed scoring model can aid in assessing the likelihood of TON in unconscious patients with periorbital fractures, facilitating timely and effective clinical decision-making in emergency settings.