RESUMO
PURPOSE: To compare the complication spectrum and rate of 23-G and 20-G vitrectomy for macular surgery. METHODS: This was a retrospective comparative analysis of 20-G and 23-G vitrectomy (introduced in 2007) for macular surgery due to macular pucker or macular hole performed between 2006 and 2010 in 61 and 59 eyes, respectively, by 2 experienced surgeons and 2 trainees. We assessed the adjusted hazard ratio for vitrectomy 23-G vs 20-G with a Cox proportional hazard model. We counted retinal detachment, vitreous hemorrhage, endophthalmitis (as early postoperative complications), or cataract progression (as late postoperative complication) as endpoint. We adjusted for indication, surgeon, retinopexy method, and the endotamponade. RESULTS: Follow-up averaged 712 days. Median time to first event was 385 days in the 23-G group and 342 days in the 20-G group. Cox proportional hazard analysis showed no significant difference between vitrectomy 23-G vs 20-G with regard to postoperative complications (hazard ratio 0.79, 95% confidence interval 0.41-1.52). The other covariates did not exert a statistically significant effect on the risk of adverse events. Looking at individual complications, retinal detachment was exclusively found after 20-G. CONCLUSIONS: In this homogenous large cohort, we did not find a statistically significant difference in rates of complications between 23-G and 20-G vitrectomy techniques for macular surgery. Trainees performed equally well as experienced surgeons.
