Impact of differences in diagnostic criteria when determining the incidence of contact lens-associated keratitis

Purpose. The purpose of this study is to examine the effect of differences in within-study and between-study diagnostic criteria in determining the incidence of contact lens-associated keratitis. Methods. We applied the sets of criteria for "microbial keratitis" as described in five previous studies to the dataset of Morgan et al., which documents 118 cases of contact lens-associated keratitis across a wide range of clinical severities. For each set of criteria, the incidence of contact lens-associated keratitis was calculated for the following five lens type/modality combinations: daily-wear rigid, daily-wear daily disposable hydrogel, daily-wear hydrogel, extended-wear hydrogel, and extended-wear silicone hydrogel. The effect of varying the clinical severity score for the differentiation of nonsevere versus severe keratitis was also examined with respect to the dataset of Morgan et al. Results. The size and location of the corneal infiltrative events identified as representing "microbial keratitis" for each of the different sets of criteria are illustrated in a series of cartograms. A key between-study difference in the incidence values calculated for the various sets of criteria relates to the categories of extended-wear hydrogel and extended-wear silicone hydrogel lenses. Specifically, the incidence of "microbial keratitis" was found to be statistically significantly greater for extended-wear hydrogel compared with extended-wear silicone hydrogel lenses when the set of criteria of Morgan et al. was applied, but not when the other sets of criteria were applied, to the dataset of Morgan et al. Increasing the threshold clinical severity criterion for differentiating between nonsevere and severe keratitis within this dataset resulted in lower incidence values; however, such changes in threshold had a minimal impact on relative risk values. Conclusions. The choice of criteria for diagnosing contact lens-associated "microbial keratitis" has a significant impact on calculations of the incidence of this condition. Copyright © 2006 American Academy of Optometry.