ObjectiveTo observe and analyze the risk factors related to vitreous re-hemorrhage (PVH) after anti-VEGF drugs combined with vitrectomy (PPV) in patients with proliferative diabetic retinopathy (PDR).MethodsRetrospective analysis study. From April 2017 to July 2018, 100 eyes of 87 PDR patients who were diagnosed in Jiaxing Eye Hospital and received anti-VEGF drugs combined with 25G PPV were included in the study. Among them, there were 44 eyes in 38 males and 56 eyes in 49 females. The age ranged from 26 to 83 years, with an average age of 57.72±8.82 years. All patients were type 2 diabetes, with an average duration of diabetes 10.84±6.03 years. All affected eyes were assisted by the same doctor with a non-contact wide-angle lens under the standard three-channel 25G PPV of the flat part of the ciliary body. Five to 7 days before the operation, intravitreal injection of ranibizumab or conbercept 0.05 ml (10 mg/ml) was performed. The incidence of PVH was observed. The age of PVH patients, duration of diabetes, vision before operation, average fasting blood glucose and average postprandial blood glucose before operation, systolic blood pressure and diastolic blood pressure before surgery, laser treatment before surgery, lens removal during operation, intraocular filling during operation, retinal laser points during operation, and fundus lesions during operation (hyperplasia film, Retinal hemorrhage, vascular occlusion, proliferative retinal traction, retinal hiatus, retinal detachment, exudation, neovascularization) were analyzed to find out the cause of PVH. Spearman bivariate correlation analysis and binary logistic regression analysis were performed on the data.ResultsOf the 100 eyes of 87 patients, PVH occurred in 17 eyes (17%). There were statistically significant differences in the number of eyes with vascular occlusion and proliferative traction during surgery in patients with and without PVH (χ2=5.741, 8.103; P＜0.05). There was no significant difference in age (t=－1.364), duration of diabetes (t=0.538), preoperative vision (t=1.897), preoperative fasting blood glucose level (t=1.938), preoperative postprandial blood glucose level (t=1.508), preoperative systolic blood pressure (t=－0.571), preoperative diastolic blood pressure (t=0.275), whether received laser treatment (χ2=2.678), the number of laser points during operation (t=0.565), whether received lens removal during operation (χ2=0.331), whether found new blood vessels during operation (χ2=2.741) and whether received intraocular filling during operation (χ2=0.060) between the patients with and without PVH (P＞0.05). Spearman's bivariate correlation analysis showed that patients with low vision, poor control of fasting blood glucose levels, vascular occlusion and proliferative retinal traction during the operation were related risk factors for PVH (rs=0.208, 0.229, 0.240, 0.285; P＜0.05). Binary logistic regression analysis showed that fundus vascular occlusion and hyperplastic retinal traction may be independent risk factors for PVH during surgery (OR=5.175, 13.915; P＜0.05).ConclusionFundus vascular occlusion and retinal traction caused by fibrovascular membrane hyperplasia in PPV may be independent risk factors for PVH in patients with PDR after anti-VEGF drugs combined with PPV.