The finding by Srinivasan et al1 in the Steroids for Corneal Ulcers Trial that topical corticosteroids do not influence outcome in the treatment of bacterial keratitis is an important advancement in our understanding. In a subgroup analysis, the authors reported that corticosteroids may have a benefit over placebo when the baseline best spectacle-corrected visual acuity is counting fingers or worse (P = .03) and in ulcers located completely centrally (P = .02). While these findings are certainly worthy of further study, it is worth considering some potential pitfalls of the analysis.2