Profile and outcomes of retinal artery occlusion: The underrealized need to expedite presentation.

Purpose: To evaluate the clinical profile of retinal artery occlusion (RAO) and impact of presentation on visual outcomes.

Design: A retrospective analysis of case files of 3070 patients with RAO was performed using electronic medical records.

Methods: Demographic data were analyzed using descriptive statistics. The differential distribution of risk factors of RAO with age was studied. Time to presentation and treatment effects were assessed using multivariate regression. Interaction plots were drawn to assess the impact of risk factors on outcomes.

Results: Central RAO was the most common type (n = 2443, 77.11%), followed by branch RAO (n = 500, 15.78%), while combined retinal vascular occlusion and cilio-RAO were rare. Most of the patients (71.40%) were male and had unilateral (96.81%) affliction. Almost half presented within the fifth (24.85%) and sixth (21.4%) decades of life. Hyperhomocysteinemia had a higher association (1.95, P = 0.0019) with younger patients (<40 years), while hypertension (3.64, P < 0.001), diabetes mellitus (DM; 4.18, P < 0.001), and coronary artery disease (CAD) (4.26, P = 0.002) were significantly commoner in older patients. CAD (5.1%) and cerebrovascular disease (0.6%) were detected after ocular presentation in some patients. Embolus, though detected rarely (1%), was associated (60%) with serious systemic disorders. Early presentation (<6 h) was associated with better visual outcomes (mean log of minimum angle of resolution 1.8 ± 1.3 vs. 2.1 ± 1.3, P = 0.032). Neovascular glaucoma (2.14% overall) was twice more common in DM (P < 0.0004) and led to further vision loss. Visual improvement occurred in 10% of patients over follow-up.

Conclusion: RAO occurs slightly earlier compared to coronary and cerebrovascular disorders and can precede their detection. Associated risk factors vary with the age of presentation. Presentation within 6 h needs facilitation for better outcomes and management.