We thank Dr Kim for his insight into our article. Our study is not as well designed as the ARCANE study, because we did not follow up sensitivity to antibiotics over time within a patient. However, measures were taken to minimize the confounders mentioned by Dr Kim. A clear standard of care had been established previously at our institution for intravitreal injections. A single application of a topical fluoroquinolone was instilled in the eye after each injection, and each patient was supplied with a new sample of topical fluoroquinolone at each visit, with written instructions that were verbalized by the patient before discharge from the clinic. Almost all previous injections (more than 95%) had been performed at our institution. It is also true that eyes in our study received intravitreal injections over different intervals; however, most were within 1 to 2 months of each other. This also was true in the ARCANE study, in which eyes received as few as 5 or as many as 13 injections within 1 year.

One strength of our study is the large number of patients (n = 104) versus the 18 in the ARCANE study. Additionally, we reported all pathogenic organisms, not just Staphylococcus and Streptococcus , isolated from the nares and conjunctiva, with their sensitivity profiles. Our study shows that more than 30% of patients with no previous injections had at least 1 conjunctival organism resistant to fluoroquinolones; the ARCANE study had similar results. This prevalence also is apparent in the nares. This finding, consistent in both studies, is an important contribution to the literature.

We, like Dr Kim, were surprised that we did not find a large increase in the prevalence of resistance among patients with increased exposure to fluoroquinolones. Patients with at least 1 fluoroquinolone-resistant organism in our study had 1.4 more injections than those without a resistant organism, and the 95% confidence interval indicates that resistant patients probably had fewer than 4 more injections than patients without a resistant organism. It would be interesting to see a valid analysis of the longitudinal data in the well-designed ARCANE study. The appropriate analysis would be to determine how many isolates that were sensitive to an antibiotic at baseline became resistant after injections. The analysis they present includes all resistant isolates after all injections in the numerator (including isolates resistant at baseline and isolates from subsequent cultures in patients who had resistant organisms in previous cultures). For example, the numerator should be the number of Staphylococcus epidermidis isolates in these eyes that became resistant during follow-up; the denominator to determine the percent of S. epidermidis isolates that became resistant must be less than 18 (assuming no more than 1 strain of S. epidermidis in each eye), not 77. With the correct denominators and numerators, the ARCANE study, like ours, may show no effect or a small effect of repeated injections on resistance. A study designed like the ARCANE study, with more patients and a more appropriate analysis, would be valuable. The data to calculate a sample size can be gleaned from the literature, including the ARCANE study and ours.