RESUMO
BACKGROUND: The aim of this study was to compare the effectiveness of pars plana vitrectomy (PPV) and removal of the internal limiting membrane (ILM) with three, monthly, intravitreal bevacizumab (IVB) injections for refractory diabetic macular edema. MATERIALS AND METHODS: This was a prospective, randomized, comparative, interventional study. Forty-four patients were enrolled and randomized in two groups. Twenty-two eyes enrolled in Group I received three IVB injections at monthly interval. Twenty-two eyes were enrolled in Group II which underwent PPV with ILM removal. The primary outcomes measured were: (1) Best corrected logMAR visual acuity (BCVA) using Snellen's visual acuity chart. (2) Central macular thickness (CMT) on optical coherence tomography. The secondary outcome measures were: Complication rates like (1) progression of lens opacities, (2) high intraocular pressure needing further treatment/procedure, (3) development of vitreous hemorrhage related to the procedure employed, (4) retinal detachment and (5) severe inflammation/endophthalmitis. RESULTS: In Group I (IVB): 3 (13.6%) eyes showed no change in BCVA; 3 (13.6%) eyes reported decrease in BCVA and 16 (72.8%) eyes showed improvement in BCVA; (P = 0.0181). In Group II (PPV): 4 (18.2%) eyes showed no change in BCVA; 5 (22.7%) eyes showed decrease and 13 (59.1%) eyes showed improvement in BCVA (P = 0.0281). Mean decrease in CMT in IVB group was 108.45 µ, whereas mean decrease in CMT in PPV group was 161.36 µ. No major complications were seen in either group. CONCLUSION: Posttreatment decrease in CMT was more in PPV group and vision improvement more in IVB group. However, no statistically significant difference between the two methods was found.