Objective  To analyze the risk factors of postoperative vitreous hemorrhage (PVH) after pars plana vitrectomy (PPV) for vitreous hemorrhage (VH) secondary to retinal vein occlusion (RVO). Methods  A retrospective case-control study. A total of 195 RVO patients (195 eyes) with VH were first treated with PPV from November 2015 to December 2021 were included in this study. There were 102 males (102 eyes) and 93 females (93 eyes), with an age of (62.93±9.78) years. The patients were divided into PVH group (17 patients, 8.72%) and non-PVH group (178 patients, 91.28%) according to the occurrence of PVH. The time of occurrence of PVH was (140.33±130.85) days after PPV. All eyes were performed 23G or 25G systematic PPV by the same doctor. During the operation, different types of intraocular tamponade and intravitreal injection of anti-vascular endothelial growth factor or triamcinolone acetonide after operation were selected according to the severity of retinopathy. The follow-up time was (9.45±6.68) months. The baseline systemic parameters, ocular parameters and intraoperative parameters affecting the occurrence of PVH were analyzed. Baseline systemic parameters included sex, age, diabetes mellitus and hypertension; ocular parameters included RVO type, lens status, VH course, preoperative best corrected visual acuity and intraocular pressure; intraoperative parameters included cataract phacoemulsification, removal of internal limiting membrane, type of intraocular tamponade, type of intravitreal injection drug at the end of operation, etc. Kaplan-Meier survival analysis, and Cox univariate and multivariate regression analysis were performed to analyze the risk factors of PVH after PPV in RVO with VH patients. Results  In PVH group, the number of patients with diabetes was more than that in the non-PVH group, and the course of diabetes was longer, and differences were statistically significant. There were significant differences in RVO type, lens status and type of intraocular tamponade. Univariate Cox regression analysis showed that the combination with diabetes [odds ratio (OR)=2.724, 95% confidence interval (CI) 1.006-7.374, P=0.049], duration of diabetes (OR=1.071, 95%CI 1.013-1.134, P=0.016), central retinal vein occlusion (OR=4.387, 95%CI 1.421-13.546, P=0.010), intraocular lens (OR=3.493, 95%CI 1.229-9.925, P=0.019), and intraocular gas tamponade (OR=3.640, 95%CI 1.365-9.702, P=0.010) were associated with PVH. Multivariate Cox regression analysis showed that intraocular gas tamponade was independent risk factor for PVH. Conclusion  Intraocular gas tamponade can increase the risk of PVH after PPV in patients with VH secondary to RVO.

Citation： Zhang Luyuan, Zhang Yan, Li Xiaorong. Risk factors of postoperative vitreous hemorrhage after pars plana vitrectomy for vitreous hemorrhage secondary to retinal vein occlusion. Chinese Journal of Ocular Fundus Diseases, 2023, 39(5): 368-374. doi: 10.3760/cma.j.cn511434-20221115-00605 Copy