Congratulations to Drs. Matthew Remz, Michael Okun and David Vaillancourt on the publication of “On the Fence: Examining Disagreement in Clinical Diagnoses of Atypical Parkinsonism,” which appears in the April issue of Neurology.
Abstract
Objective:
To examine the clinical diagnosis of atypical parkinsonism, leveraging cases in a large multicenter cohort where there was disagreement between experts.
Background:
Atypical parkinsonism is a diagnostic challenge; expert opinion frequently differs from pathological diagnosis. Studies of biomarkers often rely on a consensus diagnosis between experts as the reference gold standard. We aimed to study cases without diagnostic consensus.
Design/Methods:
We examined cases from a 21-site multicenter study focused on diffusion MRI biomarkers for differentiating parkinsonism (AID-P) where there was disagreement at the initial study visit between expert movement disorders neurologists on diagnosis. Experts comprised the referring site clinician and two blinded clinicians who reviewed videos and clinical data. We used Chi-Square tests to compare ratios of MSA:PSP diagnoses in cases with one dissenting opinion to cases with consensus between all raters.
Results:
Of 283 cases examined, 247 had diagnostic consensus, 15 had consensus from video raters but site clinician dissent, 19 had dissent of 1 video rater, and 2 had no agreement across all raters. When defining diagnosis by agreement between 2 of 3 experts, we compared ratios of MSA to PSP. Cases with 1 dissenting video rater did not vary in distribution of diagnoses from cases with full consensus (c2=1.89, p=0.1619), however cases with site clinician dissent differed (c2=10.19, p=0.0014). PSP diagnosis was more likely when the dissenting party was the site clinician (Likelihood ratio=1.99) and half as likely when the dissenting opinion was a video rater (Likelihood ratio=0.41). Multiple clinical rating scales (UMSARS, PSP-RS, MOCA) disclosed no significant differences between groups.
Conclusions:
In a large multicenter cohort of atypical parkinsonism, 12.7% of cases revealed diagnostic disagreement. Site clinician disagreement with PSP diagnosis was overrepresented. For clinical trials, it is imperative that the groups studied are homogeneous, and this study underscores the importance of using multiple raters.