Recurrent vitreous hemorrhage following pars plana vitrectomy (PPV) for proliferative diabetic retinopathy remains a significant complication with reported incidence ranging from 11 to 75%. Early and late recurrences are associated with various factors, including residual blood, fibrovascular tissue, and neovascularization. Despite attempts to reduce this complication with therapies like anti-fibrinolytic agents, gas tamponade, and peripheral cryotherapy, the outcomes remain unsatisfactory.

Recurrent vitreous hemorrhage following pars plana vitrectomy (PPV) for proliferative diabetic retinopathy remains a significant challenge with high incidence and negative impact on visual outcomes. Despite various treatments, intravitreal bevacizumab (IVB) has shown potential in reducing post-operative vitreous hemorrhage (PO-VH). This study aims to determine the efficacy and safety of intraoperative IVB in preventing PO-VH in patients undergoing vitrectomy for diabetic vitreous hemorrhage. This prospective, randomized controlled trial compared patients receiving intraoperative IVB to a control group. The primary outcome was PO-VH incidence at one month, with secondary outcomes including visual acuity change, time to VH clearance, and safety. Outcome assessors were masked to the study treatment. Data on demographics, medical history, visual acuity, VH grading, intraoperative findings, and postoperative complications will be collected. Statistical analysis compared PO-VH incidence and other outcomes between groups. This study will provide valuable evidence on the efficacy and safety of intraoperative IVB in preventing PO-VH, contributing to improved clinical practice and patient outcomes.

Inclusion Criteria:

• Patients with spontaneous, non-clearing vitreous hemorrhage for at least one month as the initial vitrectomy indication for diabetic retinopathy.

Exclusion Criteria:

• Patients with tractional retinal detachment (ART)
• Patients with severe fibrovascular traction,
• Patients with intravitreal gas or silicone oil tamponade at the end of surgery
• Patients with other eye diseases besides diabetic vitreous hemorrhage
• Patients with history of intravitreal bevacizumab injection in the last 3 months before surgery
• Patients with other health condition includes uncontrolled hypertension, a history of coagulopathy
• Patients with inability to attend follow-up examinations for at least 1 month
• Patients taking anti-aggregation and anti-platelet medications with abnormal bleeding time (BT) and clotting time (CT)

While it was not possible to mask participants and surgeons to the treatment allocation due to the nature of the surgical intervention, efforts were made to mask the outcome assessors. This involved:

Centralized Data Collection: All data related to patient outcomes (visual acuity, VH grading, complications) were collected on standardized forms and sent to a central data management center.

Coded Data Entry: Patient identification numbers were used instead of names to prevent bias.

Masked Data Analysis: The statistician conducting the analysis was blinded to group allocation.