ABSTRACT
Overdiagnosis of Parkinson's disease (PD) is suggested by specialist review of community diagnosis, and in postmortem studies. In specialist centers 4 to 15% of patients entered into clinical trials as early PD do not have functional imaging support for a PD diagnosis. In a European multicenter, prospective, longitudinal study, we compared clinical diagnosis with functional SPECT imaging using [123I]FP-CIT (DaTSCAN, GE Healthcare). Repeat observations were performed over 3 years in patients with tremor and/or parkinsonism in whom there was initial diagnostic uncertainty between degenerative parkinsonism and nondegenerative tremor disorders. Video-recording of clinical features was scored independently of functional imaging results by two blinded clinicians at 36 months (= gold standard clinical diagnosis). Three readers, unaware of the clinical diagnosis, classified the images as normal or abnormal by visual inspection. The main endpoint was the sensitivity and specificity of SPECT imaging at baseline compared with the gold standard. In 99 patients completing the three serial assessments, on-site clinical diagnosis overdiagnosed degenerative parkinsonism at baseline in diagnostically uncertain cases compared with the gold standard clinical diagnosis (at 36 months), the latter giving a sensitivity of 93% and specificity of 46%. The corresponding baseline [123I]FP-CIT SPECT results showed a mean sensitivity of 78% and a specificity of 97%. Inter-reader agreement for rating scans as normal or abnormal was high (Cohen's kappa = 0.94-0.97).
