Redefining the accuracy of clinical diagnosis of Parkinson's disease: A systematic review and meta-analysis.
Daniele Urso, Stefano Giannoni-Luza, Matteo Guidetti et al.
✦ AI-curated · Sources linked
In 30 seconds
This systematic review and meta-analysis assessed the accuracy of clinical Parkinson’s disease (PD) diagnosis against neuropathological confirmation, comparing UKPDSBB, Gelb, and MDS-PD 2015 criteria. Twelve studies including 9,659 participants were analysed. At initial assessment, pooled sensitivity was 57% and specificity was 76%, with substantial heterogeneity; at final recorded assessment, these rose to 87% and 89%, respectively. UKPDSBB had the highest final-assessment sensitivity (91%), while MDS-PD 2015 criteria had the highest specificity. The authors conclude that diagnostic accuracy depends on assessment timing and remains limited initially.
Key findings
- At initial assessment, pooled sensitivity was 57% and specificity was 76%, with substantial heterogeneity.
- At final recorded assessment, pooled sensitivity was 87% and specificity was 89%.
- At final assessment, UKPDSBB had the highest sensitivity (91%); MDS-PD 2015 criteria had the highest specificity.
- Meta-regression found specificity increased over recent decades, alongside a modest decrease in sensitivity.
Why it matters
The marked difference between initial and final diagnostic performance highlights uncertainty in early clinical classification of PD. This matters when interpreting diagnoses made at first assessment and when selecting participants for early-intervention trials. The findings also show that diagnostic criteria may balance sensitivity and specificity differently, while the evidence for comparative performance remains limited.
What to keep in mind
The review included a limited number of studies, and initial-assessment estimates showed substantial heterogeneity. The authors note that comparative findings, particularly for MDS-PD 2015 categories, require further clinicopathological validation.