Abstract
Purpose: :
To evaluate the visual outcomes of patients with diabetic vitreous hemmorhage and correlate it with potential clinical predictors such as HbA1c. Further, to look at co-morbidities of the vitreous hemmorage including recurrent hemmorhage and retinal detachments.
Methods: :
A Retrospective review of all retina cases (approx. 650) at University of Florida, Gainesville VA Medical Center between January 1, 2003 and December 31, 2006 was performed. Inclusion criteria were defined as surgeries performed for vitreous hemmorhage during the above specified time frame. Exclusion criteria included non-diabetic vitreous hemmorage and those eyes with previously performed vitrectomy as determined per patient history and documentation. A Student's paired t-test was used for statistical analysis. The following Visual Acuity Scale was used : LP-CF=1, 20/400=2, 20/200-20/80=3, 20/70-20/40=4, 20/30-20/20=1.
Results: :
A total of seventy-five cases fullfilled the above inclusion/exclusion criteria. The mean HbA1c was found to be 7.96+/- 1.53. The was an absolute correlation between HbA1c levels (those with HbA1c <8.0 vs. HbA1c>8.0) and final visual outcome (mean interval time >1yr), however this was not statistically significant amongst the subgroupings. There was a substantial decrease in visual acuity in those patients that had concomitant vitreous hemmorhage and retinal detachment (p<0.O5). On subgroup analysis there was a slightly greater percentage of recurrent vitreous hemmorage in those that were on blood thinners (coumadin, aspirin, plavix), however this did not appear to be statistically significant (P> 0.05) Vitreous hemmorhage was observed non-statistically in those with and without those with previous pan-retinal photocoagulation (PRP).
Conclusions: :
The visual outcome in diabetic vitreous hemmorhage does correlate with a patient's HbA1c level, however subgrouping analysis shows it may not be statistically relevant. Although certain patient's have favorable results following vitrectomy and panretinal photocoagulation, the diabetic vitreous hemmorage poses a significant overall morbidity. As expected, the visual outcome is worse with co-morbidities such as retinal detachment and recurrent hemmorhages. Furthermore, it appears that anticoagulation medications do not contribute to an increase in the recurrence of vitreous hemmorage in the diabetic patient.
Keywords: diabetic retinopathy • retina • vascular endothelial growth factor