Reader response: A predictive model to identify Parkinson disease from administrative claims data

Searles Nielsen et al. conducted a case-control study for predicting Parkinson disease (PD) using demographic data, ever smoking tobacco, constipation, taste/smell disturbance, REM sleep behavior disorder, and diagnosis information. They conducted a full statistical model and presented an acceptable model performance by using receiver operator characteristic area under the curve, sensitivity, specificity, and correct classification rate.¹