Author’s Disclosure Block: Keean Nanji: none; Dana Taghaddos: none; Lele Xiao: none; Justin Grad: none; Phillip Staibano: none; Tyler McKechnie: none; Michael Zoratti: none; Varun Chaudhary: Dr. Chaudhary reports acting as an advisory board member, grants and other from Novartis; acting as an advisory board member, grants and other from Bayer; grants from Allergan; acting as an advisory board member and grants from Roche; acting as an advisory board member for Janssen; acting as an advisory board member for Apellis; and acting as an advisory board member for Boehringer Ingelheim outside the submitted work. 

Abstract Body
Purpose: Cystoid macular edema (CME) is among the most common complications following cataract surgery. The purpose of this review wasto synthesize the evidence from randomized controlled trials evaluating patients receiving prophylactic treatment with nonsteroidal anti-inflammatory drugs or corticosteroid medications to determine the comparative effectiveness of each specific regimen on preventing post-operative cystoid macular edema and key patient-important outcomes following cataract surgery. Study Design: Systematic review and random effects Bayesian network meta-analysis (NMA). Methods: A comprehensive search was conducted of the MEDLINE, EMBASE and CENTRAL databases, as well as of the ClinicalTrials.gov and World Health Organization International Clinical Trials Registries from inception to February 27th,2024.The protocol was registered a priori on PROSPERO (CRD42024531150).Data was synthesized for seven pre-specified outcomes at six weeks and three months following surgery: i) change in central retinal thickness measured by optical coherence tomography (OCT), ii) best-recorded visual acuity iii) the rate of CME measured by OCT, and fluorescein angiography (FA) iv) the rate of patients experiencing clinically significant macular edema defined as the presence of CME and pre-specified thresholds for decreased visual acuity, v) patient quality of life, vi) intraocular pressure and vii) the number of patients experiencing one or more pre-specified adverse events. Bayesian network meta-analysis with random effects modelling was performed for each comparison. Pre-specified network meta-regression evaluated the impact of baseline diabetes status, previous uveitis history, and whether patients underwent femtosecond laser-assisted surgery. Results: 67 trials of 11,294 eyes were included. 16 regimens were evaluated when grouping medications by route, class, and timing (e.g. pre-op topical steroids (TS) separated from pre-op and post-op TS) and 10 regimens grouping by route and class.Topical NSAID (TN) with TS reduced the rate of cystoid macular edema (CME), CME with vision loss, and improved retinal thickness. Network meta-regression, controlling for diabetes, suggested TS + TN improved vision at three months only in patients with diabetes. TS combined with intra-op intravitreal or sub-tenon steroid injection reduced CME incidence, CME with vision loss, but not vision. Post-op TS with intravitreal steroid implant improved vision. There was no benefit to giving regimens pre-op in addition to post-op. Conclusion: Compared with TS, many regimens improved retinal thickness or CME incidence but not outcomes incorporating vision. Only TS+TN, TS with intravitreal steroids, TS with sub-tenon steroids, or TS with steroid implants improved the rate of CME with vision loss or improved vision.