The authors show convincing findings with a rigorous study design. One challenging variable in any study is assigning appropriate diagnostic criteria within the involved cohort. To identify patients with newly diagnosed epilepsy, Kalilani et al used criteria validated in a Canadian-based study by Reid et al.8 Reid et al evaluated several large insurance and claims databases, validating results with chart reviews from patients from 13 neurologists’ practices. Given the much higher prevalence of epilepsy in the neurologists’ practice as compared to the general population, there is a risk that the coding algorithms showed an inflated positive predictive value in the neurologists’ practice group. A larger concern is how the results from Kalilani et al, which included patients with Medicaid and Medicare, apply to other populations with different demographics. The authors discuss this important issue in the article.