PURPOSE: To investigate potential risk factors for development of proliferative vitreoretinopathy (PVR) post trauma and evaluate the effect of PVR on anatomical and visual outcomes in injured eyes. METHODS: Overall, 179 eyes with PVR and 221 eyes without PVR after injury were selected from the database of the Eye Injury Vitrectomy Study, a multicenter cohort study launched in 1997. Multivariate logistic regression was used to ascertain the independent risk factors for development of PVR and to evaluate the influence of PVR on anatomical and visual outcomes. RESULTS: An interval of injury and vitrectomy of more than 28 days (odds ratio, 139.25; confidence interval, 50.09-387.10), severe vitreous hemorrhage (odds ratio, 2.72; confidence interval, 1.13-6.52), and total retinal detachment (odds ratio, 12.67; confidence interval, 3.96-40.52) were important independent risk factors for PVR. One hundred and fifteen eyes (52.0%) and 49 eyes (27.4%) without and with PVR, respectively, were anatomically restored with ambulant visual acuity (≥4/200). Proliferative vitreoretinopathy, poor initial visual acuity, relative afferent pupillary defect, total retinal detachment, and retinal tear or retinal defect were unfavorable prognostic indicators. CONCLUSION: Proliferative vitreoretinopathy occurs frequently in injured eyes and is associated with poor outcomes. Its onset depends on interval of injury and vitrectomy, wound location, vitreous hemorrhage, and retinal detachment. Early vitrectomy (before 2 weeks) and aggressive therapy should be considered for specific high-risk cases.
PURPOSE: To investigate potential risk factors for development of proliferative vitreoretinopathy (PVR) post trauma and evaluate the effect of PVR on anatomical and visual outcomes in injured eyes. METHODS: Overall, 179 eyes with PVR and 221 eyes without PVR after injury were selected from the database of the Eye Injury Vitrectomy Study, a multicenter cohort study launched in 1997. Multivariate logistic regression was used to ascertain the independent risk factors for development of PVR and to evaluate the influence of PVR on anatomical and visual outcomes. RESULTS: An interval of injury and vitrectomy of more than 28 days (odds ratio, 139.25; confidence interval, 50.09-387.10), severe vitreous hemorrhage (odds ratio, 2.72; confidence interval, 1.13-6.52), and total retinal detachment (odds ratio, 12.67; confidence interval, 3.96-40.52) were important independent risk factors for PVR. One hundred and fifteen eyes (52.0%) and 49 eyes (27.4%) without and with PVR, respectively, were anatomically restored with ambulant visual acuity (≥4/200). Proliferative vitreoretinopathy, poor initial visual acuity, relative afferent pupillary defect, total retinal detachment, and retinal tear or retinal defect were unfavorable prognostic indicators. CONCLUSION:Proliferative vitreoretinopathy occurs frequently in injured eyes and is associated with poor outcomes. Its onset depends on interval of injury and vitrectomy, wound location, vitreous hemorrhage, and retinal detachment. Early vitrectomy (before 2 weeks) and aggressive therapy should be considered for specific high-risk cases.