Benefits of noninferiority studies in ophthalmology

Abstract

We have read the article by Tyson et al.1; however, we believe some discussion is required. The authors present the results in a study of the treatment of ocular inflammation and pain after cataract surgery with a sustained-release intracanalicular dexamethasone insert. Because the outcomes in contemporary cataract surgery partially depend on patient compliance, new therapeutic approaches (eg, an intracanalicular insert) represent an appealing alternative. Nevertheless, the study is not principally different from a previous trial2 and does not present significantly more information. In the study, patients were randomized after cataract surgery to receive a dexamethasone intracanalicular insert or a placebo. Other antiinflammatory treatment was not permitted preoperatively, intraoperatively, or postoperatively. This part of the study raises ethical concerns. Corticosteroids, which are routine in antiinflammatory management after cataract surgery, were not administered in the placebo arm; therefore, the patients in this group were exposed to a higher risk for adverse events.3 Thus, we should ask whether the patients were informed about these risks. Such concerns were confirmed by the results in the study, In the placebo arm, patients more commonly experienced eye pain 8 days after surgery (70.0% in placebo group versus 20.7% in treatment group), conjunctival hyperemia, and corneal edema and had a reduction in visual acuity. Moreover, the same therapeutic hypothesis could be tested in a different research model. In this case, we believe that using a noninferiority study design would be beneficial. Since the 1990s, placebo-controlled trials have become ethically controversial if a standard treatment exists. A noninferiority trial aims to show that an experimental treatment is not (much) worse than the accepted one.4 The new treatment might be safer, have better secondary endpoints, be easier to administer, or simply be less expensive. In ophthalmology, noninferiority trials are commonly applied for analyzing outcomes of age-related macular degeneration therapy and glaucoma treatment with prostaglandins. We believe that comparing the efficacy of the dexamethasone insert with the accepted topical corticosteroid treatment would have been the best option for this study. Patients in the placebo group had more ocular adverse events, and this could have been prevented with such a design. Because a mean dose compliance for topical treatment after cataract surgery might reach only 50.2%,5 the intracanalicular insert would be an interesting alternative.