Purpose
To identify prognostic indicators of postoperative visual acuity and retinal detachment among patients with open globe injuries, as well as to compare our findings to an established algorithm for predicting long-term visual outcomes from ocular trauma.
Methods
Retrospective chart review consisting of 54 eyes from 54 adult patients (38 male, 16 female) diagnosed with open globe injury between September, 2011 and March, 2014. Inclusion criteria included full assessment of Ocular Trauma Score criteria (preoperative visual acuity, mechanism of injury, reverse afferent pupillary defect, retinal detachment, and endophthalmitis), as well as data on demographics and other injury characteristics (i.e. zone of injury, lid/adnexal injury, hyphema, vitreous hemorrhage, loss of uveal tissue and traumatic cataract). Multivariate linear regression was used to assess the relationship between potential predictors and postoperative visual acuity (FVA), after adjustment for age, sex, race and preoperative visual acuity (IVA). Logistic regression was used to evaluate the relationship between potential predictors and occurrence of retinal detachment (RD).
Results
Mean participant age was 45.6 years, and mean follow-up time was 1.2 years. Blunt trauma accounted for 57.4% (31/54) of injuries. Globe perforation occurred in 14.8% (8/54) of cases, and an intraocular foreign body was present in 20.4% (11/54) of cases. The prevalence of RD was 42.6% (23/54). Mean time to RD was 9.3 days. Relative afferent pupillary defect was observed in 33.3% (18/54) of patients. Only one patient developed endophthalmitis. The base model including IVA accounted for 62.0% of the R2 in FVA. Adding information about RD significantly improved R2 to 70.6%. In comparison, the ocular trauma score accounted for 65.7% of R2 in FVA. No other injury characteristic was associated with FVA after adjustment for IVA. Among patients with 20/200 or worse FVA, 100% (27/27) had IVA of 20/200 or worse, and 70.4% (19/27) also had RD. The same base model including IVA accounted for 38.8% of the R2 in RD. Adding information on vitreous hemorrhage significantly improved R2 to 60.6%. No other injury characteristic was associated with RD after adjustment for IVA. Among patientswith RD, 95.7% (22/23) also exhibited vitreous hemorrhage.
Conclusions
IVA and occurrence of RD were the most important prognostic indicators of FVA. IVA and vitreous hemorrhage were significant predictors of RD.