Non-inferiority trials in clinical ophthalmology

Abstract

Purpose: To summarize the characteristics and methodology of non-inferiority trials in ophthalmology, aiding researchers in understanding the applications and limitations of such trials in ophthalmic diseases. Methods: PubMed, Web of Science, Embase and Scopus were searched for literature on non-inferiority randomized trials in ophthalmology published between 2000 and November 5 2023. Data on the basic characteristics were extracted and summarized. The Risk of Bias 2’s was used to assess the bias risk. Results: A total of 294 papers were included, with 77.6% of the trials conducted in the last 10 years, and more than 2/3 (72.1%) were multicenter studies, and 79.9% were registered on platforms. The majority of trials were applied in the researches of glaucoma, cataract, age macular degeneration, macular edema, dry eye, myopia, or refractive error. Non-inferiority thresholds were reported in 88.4% of the trials. Intent-to-treat analysis was the primary outcome analysis method in only 21.8% of trials, while both intent-to-treat and per-protocol analyses were used in 29.6%. Last observation carried forward method was used to address missing values in 23.5%. However, 56.5% of the articles did not report how missing values were handled, leaving uncertainty regarding whether missing data was considered in the analysis. About 20.7% of the studies were at high risk of bias, mainly due to outcome measures and missing value treatments. Conclusion: Non-inferiority trials are commonly used in ophthalmologic research to assess the effectiveness, safety, cost-effectiveness of treatments or surgical methods, but the quality of implementation and reporting needs to be improved.