JMIR Publications

ABSTRACT

Background:

Traditional in-person neuropsychological tests for Parkinson’s disease (PD) lack accessibility, scalability, and PD-specificity. Mobility impairments hinder access to in-person assessments, and long waiting times for expert assessments limit scalability. Common tools for cognitive screening, such as the MoCA, are generic and not specific to PD. Our goal was to address these challenges by leveraging the internet.

Objective:

We aimed to develop a sensitive tool to detect cognitive impairments in early to mid-stage PD in an accessible and scalable manner.

Methods:

We developed the Online Rapid Cognitive Assessment for PD (ORCA-PD), a brief (~15 min), fully self-administered tool, tailored to detect PD-specific impairments remotely. In a cross-sectional study, 112 participants from diverse geographical locations completed the MoCA, MDS-UPDRS-III, ORCA-PD, and a usability questionnaire. Task selection for ORCA-PD was guided by meta-analyses and comprehensive reviews, demonstrating medium-high sensitivity to PD cognitive impairments.

Results:

Participants from 30+ locations showed a 93% completion rate with median usability ratings of 4 to 5 out of 5, indicating strong usability across a diverse sample. ORCA-PD score significantly correlated with MoCA score (ρ=0.45, P<.001), supporting convergent validity. ORCA-PD score also followed expected significant trends: neurotypical controls>PD (β=4.18, P=.048), and PD-non-MCI>PD-MCI (β=7.22, P=.046), confirming two discriminative abilities. A machine-learning classifier yielded medium-to-high classification performance (AUC = 0.79) on an independent dataset, corresponding to 29% predictive power when using ORCA-PD predictors alone.

Conclusions:

ORCA-PD demonstrated usability, convergent validity, discriminability, and predictive power. ORCA-PD offers an accessible, scalable, and PD-sensitive cognitive screening test that could complement traditional in-person, supervised tools.