Objective: Eliminate intra- and inter-rater variability by using digital means for rating scales and making them accessible for all.

Background: Conventional clinician-rated scales like UPDRS and Hoehn & Yahr (H&Y) are foundational for tracking Parkinson’s disease progression but suffer from inconsistent item interpretation and low inter-/intra-rater reliability, with Intraclass Correlation Coefficient (ICC) often reported as less than 0.70. This lack of standardization compromises precision, particularly in early PD in which subtle (low score) signs on ratings predominate. Clinic-based administration further limits access for patients with mobility challenges, hindering repeated, real-world monitoring during clinical trials and monitoring medication and DBS changes.

Method: Mon4t addresses these challenges by delivering precise, objective readings of individual PD symptoms (e.g., tremor amplitude, walking speed, finger tapping rate) directly from smartphone sensors – in this case, the accelerometer, gyroscope, and touchscreen – eliminating subjective rater interpretation. We sampled 100 patients using various digital tests, extrapolated UPDRS and H&Y scores, and compared between the two.

Results: In a cohort of 100 PD patients spanning H&Y stages 1-4, Mon4t measured digital scores for key motor signs including gait, tremor, and bradykinesia. Test–retest reliability reached ICC=0.89, markedly exceeding clinician inter-rater ICCs of 0.65–0.70 reported in prior UPDRS validation studies. Digital endpoints correlated well with gold-standard MDS-UPDRS Part III as well as H&Y in total scores (r=0.81, p<0.001; r=0.84, p<0.001) and individual items (e.g., r=0.83 for finger tapping 3.7).

Conclusion: Digital markers collected by a standard smartphone can be used not only to provide accurate quantification of each symptom, but also to provide an accurate overall rating.

« Back to 2026 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/digitizing-pd-rating-scales-increases-reliability-and-accessibility/