Natural History and Surgical Timing for Idiopathic Epiretinal Membrane.

PURPOSE: This study assessed the prevalence of progression to surgery on idiopathic epiretinal membrane (iERM) and compared the results with those of immediate surgery. DESIGN: Retrospective study with nested case-control comparison. SUBJECTS: Patients with mild iERM at initial presentation METHODS: A retrospective study with a nested case-control study was conducted to compare the outcomes of those who later progressed to surgery ("deferred surgery") with those of a control group that underwent surgery within 6 months of initial presentation ("immediate surgery"). MAIN OUTCOME MEASURES: The prevalence of performing vitrectomy with membrane peeling after initial deferral, features associated with deferred surgery, and the visual acuity (VA) outcomes of surgery compared with those with immediate surgery. RESULTS: A total of 413 patients were included in the base study cohort: 369 did not undergo iERM peeling, whereas 44 (10.7%) underwent deferred surgery at a mean duration of 18.1 months after initial presentation. The factor most associated with later progression to surgery, determined using a multivariate analysis, was symptoms at initial presentation (odds ratio [95% confidence interval], 8.75 [3.80-20.15]; P < 0.0001). The immediate surgical group (controls) had a poorer logarithm of the minimum angle of resolution (logMAR) best-corrected visual acuity (BCVA) at presentation compared with the deferred surgery group (cases), but this decreased in the latter group to about the same preoperative level immediately before the deferred surgery (typically why they underwent surgery). The final logMAR BCVA was similar in the deferred and immediate surgical groups when only pseudophakic cases were analyzed. An analysis of the change from preoperative BCVA (instead of initial) to final BCVA showed no difference in the magnitude of change between the pseudophakic subgroups of the deferred and immediate surgery groups (median [interquartile range] preoperative VA of 0.52 [0.40-0.60] in the deferred surgery group and 0.44 [0.39-0.60] in the immediate surgery group [P = 0.7256]). Within the deferred and immediate surgery groups, the median change from preoperative BCVA to final BCVA in the pseudophakic subgroups was 0.22 (P = 0.0082) in the former and 0.21 (P = 0.0001) in the latter. CONCLUSIONS: The vast majority of eyes with iERM remain stable after initial presentation. The deferral of surgery in the minority of patients who progress does not have a disadvantageous final outcome compared with immediate surgery.