PROGNOSTICATORS FOR VISUAL OUTCOME IN SARCOID UVEITIS

Purpose: The purpose of the study is to delineate the visual prognosti cators in sarcoid-associated uveitis given the current standards of ca re. Methods: The records of 60 patients with sarcoid-associated uveiti s who were observed for at least 6 months were studied retrospectively . Multivariate regression models using the generalized estimating equa tions approach to adjust for the correlation between fellow eyes were applied to determine disease, patient, and treatment characteristics t hat altered the odds of visual rehabilitation. Results: One hundred tw elve eyes of 43 women and 17 men who met the inclusion criteria were i dentified. Seventy-seven percent of patients were white, 15% black, an d 8% of Hispanic origin. Uveitis developed in the patients at a mean a ge of 42 (range, 4-82) years. Of the 112 affected eyes, 81% had granul omatous and 15% nongranulomatous uveitis. Most patients (66%) had ante rior or intermediate uveitis alone. Ninety-one percent had chronically smoldering disease; another 7% had recurrent flares, and only 1 patie nt had a monophasic acute course to her uveitis. Vision-threatening co mplications developed in many patients, including 58% in whom cystoid macular edema developed and 25% in whom media opacities developed, req uiring cataract surgery or vitrectomy or both. Overall, 34% of treated eyes and 51% of patients had final visual acuities that were superior to their acuities at presentation. The factors most significantly ass ociated with a final visual acuity of worse than 20/40 after controlli ng for potential confounders were as follows: delay in presentation to a subspecialist (odds ratio [OR] = 2.94, P = 0.05), total duration of uveitis (OR = 1.04, P = 0.09), development of cystoid macular edema ( OR = 0.37, P = 0.07) or glaucoma (OR = 4.54, P = 0.02), presence of in termediate (OR = 5.00, P = 0.01) or posterior uveitis (OR = 8.33, P = 0.04), and systemic steroid use (OR = 0.30, P = 0.03) were the paramet ers most strongly correlated with a lack of visual acuity improvement. Delay in presentation to a subspecialist (OR = 20.00, P = 0.01), deve lopment of glaucoma (OR = 50.00, P = 0.005), presence of intermediate (OR = 25.00, P = 0.02) or posterior uveitis (OR = 50.00, P = 0.02), bl ack race (OR = 11.11, P = 0.02), (log) visual acuity at presentation ( OR = 0.05, P = 0.0001), and use of systemic steroids (OR = 0.07, P = 0 .02). Conclusion: Multivariate outcomes analysis, particularly after c orrecting for the correlation between fellow eyes, is a useful analyti c tool for optimization of standards of care and for disease risk stra tification to aid both physicians and patients.