Abstract
Purpose: :
To determine vision outcomes in eyes with open globe injury following prior penetrating ocular surgery, and to compare outcomes to those predicted by two ocular trauma models, the Ocular Trauma Score (OTS) and the Wilmer open globe prognostic model.
Methods: :
Retrospective chart review of all patients with open globe injuries and history of penetrating ocular surgery in the traumatized eye who presented to Harborview Medical Center, Seattle, WA between August 1, 1996 and December 31, 2006. The OTS was calculated and categorized as published by Kuhn (2002). The Wilmer Eye Institute’s "no vision outcome decision tree" and "profound vision loss decision tree" published by Schmidt (2008) were used to stratify the study eyes. Eyes lacking necessary data to fit both models were excluded. Observed and published outcomes were compared using Chi-square or Fisher’s exact test.
Results: :
Univariate analysis: Fifty-two eyes met the criteria for study. The mean age was 66 years, 26 (50%) were female. Prior surgeries included: 21 (40%) extracapsular cataract extraction (ECCE), 20 (39%) penetrating keratoplasty, 5 (10%) open globe repair, and 6 (12%) other procedures. The mean follow-up was 316 days. At last visit, 20 (39%) had no vision survival (NLP, enucleation, or evisceration). No vision survival was significantly associated (p<0.05) with presence of worse initial visual acuity, prior ECCE, prior open globe repair, rAPD, vitreous hemorrhage, or uveal prolapse.OTS comparison: Of 39 eyes in OTS Category 1, significant differences in outcomes (p<0.05) were seen in the NLP (51% observed vs 74% published) and LP/HM (38% vs 15%) groups. Of 12 eyes in Category 2, significant differences were seen in the NLP (0% vs 27%) and LP/HM (75% vs 26%) groups. No other differences were demonstrated in these two categories. One eye was in Category 3; no eyes were in Category 4 or 5.Wilmer model comparison: In the "no vision outcome decision tree," there was statistically significant difference (p<0.05) between observed and published vision survival at the third decision point (presence of lid laceration with LP vision and rAPD), but not at the other decision points. In the "profound vision loss decision tree," there was no statistically significant difference in vision survival at any decision point.
Conclusions: :
Open globe injury in eyes with prior penetrating ocular surgery portends poor vision survival. The results of this study suggest that vision survival may be better predicted by the Wilmer model than the Ocular Trauma Score in this subset of patients.
Keywords: trauma • clinical (human) or epidemiologic studies: outcomes/complications • clinical (human) or epidemiologic studies: risk factor assessment