Objective: We evaluated clinical validity of smartphone-administered active assessments and passive measures of motor function in early Parkinson’s Disease (PD) designed for in-home remote use.

Background: Using baseline data from an ongoing study, we  evaluated test-retest reliability, responsiveness to symptomatic treatment (ST) medication, and criterion validity determined by associations with aligned items of MDS-UPDRS administered in-clinic.

Method: During the evaluated baseline period, 82 participants performed 12 active assessments daily and used a wrist-worn digital health technology continually for one week.  Test-retest reliability of derived measures was evaluated using a linear mixed effects model (LME).  Responsiveness to symptomatic treatment was assessed in the 62 participants on ST medications using LMEs applied to each measure scaled by its minimal detectable change.  Associations between MDS-UPDRS parts 2 and 3 items and summaries of a week of digital measure collection were assessed using a Kruskal-Wallis test.  A Benjamini-Hochberg correction was applied to each set of tests using a false discovery rate of 0.05.

Results: 92% of 6852 scheduled active assessments were completed and analyzed.  Twenty-nine measures from 9 assessments showed good-to-excellent reliability (intraclass correlation coefficient > 0.75) and 27 were responsive to ST medication.  We found postural tremor was more responsive to ST medication than resting tremor, and one-finger tap count and speed were more responsive than two-finger.  Twelve measures (11 active and 1 passive) were significantly associated with an aligned MDS-UPDRS item, including at least one from each of the following assessments: passive gait, two-finger tapping, postural & resting tremor, and pronation/supination (Fig 1).  Four measures showed clinical validity across all 3 domains (reliability, responsiveness to ST medication, and criterion validity) and another 15 across 2 domains.

Conclusion: All active assessments resulted in measures that were both reliable and exhibited a significant association with either ST medication or MDS-UPDRS scores, suggesting clinical utility in different contexts of use.  Future longitudinal analyses will allow identification of measures that are additionally sensitive to disease progression, supporting identification of fit-for-purpose digital measures in interventional trials.

Selected digital measures vs clinician assessment

« Back to 2025 International Congress

MDS Abstracts - https://www.mdsabstracts.org/abstract/clinical-validity-of-a-digital-toolkit-combining-active-and-passive-assessments-in-early-parkinsons-disease/