Reply to Farioli and Kriebel

[[abstract]]To the Editor—Drs Farioli and Kriebel requested the incidence of rhegmatogenous retinal detachment (RRD) in the database we used, which was 11.7 per 100 000 person-years, similar to the previous study [1]. The incidence of RRD in selected patients receiving fluoroquinolones is higher than that in the general population because many of our patients had underlying diseases. For example, the prevalence of diabetes in the fluoroquinolone cohort is 17.9%, which is much higher than in the general population (3.15%–4.22%) [2]. In addition, these patients visited physicians more frequently, which could also increase the incidence of RRD (surveillance bias). That is why patients prescribed amoxicillin were selected as a control cohort, which was matched on possible confounding factors described in the literature [3] and propensity score. After matching, the control cohort therefore had a higher incidence of RRD than the general population