Abstract
Purpose :
Reports about the prevalence of keratoconus vary widely, not only due to ethnicity but also due to study design and setting. Therefore, we aimed to estimate the prevalence of keratoconus in a population-based study and to assess the utility of BAD-D (final D of the Belin-Ambrosio Display) in this context.
Methods :
The Rotterdam Study is a prospective population-based cohort study including persons aged 45 years and over. A cross-sectional analysis was performed, encompassing all participants who underwent at least one Pentacam scan of acceptable quality. Diagnosis of keratoconus was based on Scheimpflug tomography: a BAD-D of ≥2.6 in either eye with no history of corneal or refractive surgery was considered a probable case of keratoconus. The ocular history and Scheimpflug tomography outcomes of those probable cases underwent an independent point-based assessment by two corneal surgeons to exclude false-positives and keep definite keratoconus cases only. Inter-rater reliability was assessed and the role of possible risk factors such as asthma, hay fever, diabetes and itchy skin diseases was also investigated by performing logistic regression adjusted for age and sex.
Results :
A total of 3633 participants underwent scans. Of these, 2732 participants (75%; mean age±standard deviation: 59±12 years, 56% females, 94% European ancestry) had Scheimpflug tomography scans of acceptable quality. Eight participants (mean age±standard deviation 55±8 years, 38% females, 75% Europeans) were classified as definite KC, and only four (50%) of the confirmed patients reported being aware of the diagnosis. The prevalence of keratoconus in this cohort is thus 0.29%. This is a fairly conservative estimate as all suspects and forme fruste keratoconus cases were excluded in the analysis. Interestingly, a BAD-D score of ≥2.6 performed poorly as a standalone marker for keratoconus, with an alarming false positive rate of at least 72% among those who haven’t had corneal or refractive surgery before.
Conclusions :
We hereby confirm, in a population-based setting, the previous findings of Godefrooij et al. who reported relatively high keratoconus prevalence based on Dutch insurance data. Furthermore, we underline that some patients seem unaware of having the condition. The study also highlights shortcomings in using elevated BAD-D readings to assess the prevalence of keratoconus as false positive rates my be higher than previously reported.
This abstract was presented at the 2023 ARVO Annual Meeting, held in New Orleans, LA, April 23-27, 2023.