Long-term results, prognostic factors and cataract surgery after diabetic vitrectomy: a 10-year follow-up study

To report long-term results, prognostic factors and cataract surgery after diabetic vitrectomy.

Retrospective review of patient files from a large diabetes centre between 1996 and 2010. Surgical history was obtained from the Danish National Patient Register. Follow-up intervals were 3 months and 1, 3, 5 and 10 years after surgery.

In total, 167 patients had diabetic vitrectomy indicated for non-clearing vitreous haemorrhage (47%) and tractional retinal detachment (53%). The proportion of patients with visual acuity ≥0.3 increased from 29% before surgery to 60% after 3 months (p < 0.001, chi-square test). Median visual acuity increased from 0.06 before surgery to 0.3 after 3 months (p < 0.001, paired signed-rank test) and 0.4 after 1 year (p = 0.009) before stabilizing. No significant long-term prognostic factors were identified for non-clearing vitreous haemorrhage patients. For tractional retinal detachment patients, use of silicone oil was associated with low vision (visual acuity<0.3) after 3 months and 1, 3 and 5 years (all odds ratios >4 and p-values ≤ 0.03, logistic regression). Of the 134 patients who were phakic after surgery, 43% and 29% were phakic after 5 and 10 years, respectively. Use of silicone oil increased the risk of cataract surgery (p = 0.009, log-rank test).

Most diabetic vitrectomy patients stand to gain visual acuity ≥0.3 after surgery and a stable long-term visual acuity after 1 year. The only consistent long-term predictor of low vision after surgery is use of silicone oil for endotamponade. About 2/3 of phakic patients will subsequently have cataract surgery the first 10 years after diabetic vitrectomy.