Editors' note: A predictive model to identify Parkinson disease from administrative claims data

Abstract

In “A predictive model to identify Parkinson disease from administrative claims data,” the authors used demographic data, tobacco history, and diagnoses to develop a predictive model to identify Parkinson disease (PD) in the prodromal period. Dr. Kawada points out the difference in the adjusted odds ratio of having ever smoked tobacco in patients with PD in this study (0.70) with relative risk in a previous meta-analysis (0.44). He proposes looking at other forms of nicotine intake to better understand this association. Dr. Kawada also suggests looking at the effects of human leukocyte antigen complex and diet on future predictive models of PD. Authors Nielsen and Racette explain their methodology for estimating the probability of ever smoking in their data set and its limitations. In “A predictive model to identify Parkinson disease from administrative claims data,” the authors used demographic data, tobacco history, and diagnoses to develop a predictive model to identify Parkinson disease (PD) in the prodromal period. Dr. Kawada points out the difference in the adjusted odds ratio of having ever smoked tobacco in patients with PD in this study (0.70) with relative risk in a previous meta-analysis (0.44).