PII-003 - COMPLIANCE, USABILITY, RELIABILITY, AND CONSTRUCT VALIDITY OF SELF-ADMINISTERED DIGITAL MEASURES DERIVED FROM MDS-UPDRS PARTS 2 AND 3 IN AN OBSERVATIONAL STUDY TO EVALUATE PARKINSON'S DISEASE (PD) PROGRESSION: INTERIM ANALYSIS.

J. Lavine¹, J. Bakker², A. Scotina¹, S. Evans¹, G. Stephenson¹, M. Dockendorf³, R. Ellis¹, O. Harari⁴, D. Hurry¹, S. Lee³, E. Mangin³, J. Ren³, P. Schmidt⁵, M. Wipperman⁴, R. Rodriguez⁶, P. LeWitt⁷, L. Omberg¹, J. Wagner⁸; ¹Koneksa Health, New York, NY, USA, ²Koneksa Health, New York, NY, YSA, ³Merck & Co., Inc, Kenilworth, NJ, USA, ⁴Regeneron Pharmaceuticals, Inc., Tarrytown, NY, USA, ⁵Rho, Durham, NC, USA, ⁶Neurology One, Orlando, FL, USA, ⁷Wayne State University, School of Medicine, Detroit, MI, USA, ⁸Koneksa Health, New York, New York, USA.

Background: To evaluate the applicability of at-home smartphone-administered assessments of motor function and speech in participants with early Parkinson’s disease (PD), we evaluated compliance, usability of the assessments, reliability, and construct validity of the derived measures. In collaboration with pharmaceutical partners, we launched LEARNS (NCT06219629), a 12-month observational study to longitudinally evaluate outcome assessments comparable to MDS-UPDRS Parts 2 and 3 items, amongst up to 70 participants with PD.

Methods: Smartphone-based assessments are completed weekly over the course of 12-months, in both the on- and off-state where applicable, as well as during one-week bursts of daily assessments at baseline and 90-day intervals. We calculated compliance as a percentage of expected assessments completed to date, assessed usability with a questionnaire following one week of use, and calculated test-retest reliability for 36 derived digital motor and speech measures within bursts. These measures were compared to aligned MDS-UPDRS subparts (for example, Gait Freeze Index with Parts 2.13 and 3.11) completed in-clinic on Days 1 and 30. This interim analysis includes up to two months of data from 22 participants.

Results: Of completed bursts, on average 95% of assessments were completed, with slightly higher compliance in off-state versus on-state (98% versus 92%). Most participants found the smartphone easy and comfortable to use, with clear instructions (>75% responded with “Agree” or “Strongly Agree” to all questions). Of the 36 measures, 26 showed excellent or good reliability (17 with intraclass correlation coefficient [ICC] >0.9 and nine with ICC >0.75). Comparisons between individual items on MDS-UPDRS Parts 2 and 3 and aligned digital measures indicated association with severity for measures of speech, bradykinesia, gait, and tremor (see Figure).

Conclusion: The smartphone-based assessments demonstrated high compliance and usability, and digital motor and speech measures derived from these assessments demonstrated high test-retest reliability and construct validity. These interim results suggest strong potential for clinical validity and applicability of at-home digital measures in PD clinical studies. Longitudinal evaluation in a larger sample is planned.