Why Digital Endpoint Design Should Start with Patients, Not Sensors

Modern digital health technologies can measure a great deal in Parkinson’s disease; gait, balance, postural tremor, finger tapping, and continuous physical activity across weeks of free-living data. But measurement capability is not the same as measurement relevance. A study published in Frontiers in Neurology, co-authored by Koneksa Health’s Robert Ellis alongside Sanofi researchers, works the problem in the opposite direction from most digital endpoint development: before designing the toolkit, ask clinicians and patients which aspects of PD actually matter.

What the Study Did

  • Three inputs, one question: The team combined a literature review of qualitative PD research, an advisory panel of eight clinical PD experts, and an online survey of 202 people living with PD across 39 US states, 189 with idiopathic PD and 13 with GBA1-associated PD. Each input answered the same question: which everyday impacts of Parkinson’s are most meaningful to improve?
  • Where clinicians and patients agreed: Both groups converged on a familiar motor core; tremor, rigidity and stiffness, bradykinesia, global motor function, hand difficulties, and problems with mobility and balance. Patients ranked moving around as quickly as they wanted, maintaining balance, and taking part in outdoor active leisure among the impacts they most wanted to see improve.
  • Where they diverged: Depression and fatigue were rated highly meaningful by patients but only moderately so by clinicians. A gap the authors connect to the well-documented underdiagnosis of non-motor symptoms in routine PD care. Without the patient survey, those impacts would have been ranked by clinical judgment alone.
  • From priorities to instrument: The highly meaningful aspects were carried into the design of the Parkinson’s Disease Functional Impacts Digital Instrument (PD-FIDI): app-based ePRO items adapted from the MDS-UPDRS Part II, a dyskinesia assessment adapted from the MDS-UDysRS, active functional assessments for gait, balance, postural hand tremor, pronation/supination and finger tapping, and continuous wrist-worn actigraphy.
  • What didn’t make the cut, and why: Not every meaningful impact became a measure. Depression was left out over concerns about the validity of longitudinal self-report scales. Cognitive impacts were impractical to assess within the instrument. Sit-to-stand balance tasks were excluded on fall-risk grounds. The authors are explicit that fatigue, though patient-prioritized, may require alternative methods.

Why the Exclusions Are as Instructive as the Inclusions

It would have been easy to publish a paper that simply reported what patients said and implied the instrument captured it. This one does something more useful for sponsors: it names the three filters; technically viable, operationally feasible, and safe, that every candidate impact had to pass, and documents the meaningful impacts that failed them.

That is the realistic shape of digital endpoint development. Patient input defines the target; technical and operational constraints define what can be measured well enough to support a regulatory or clinical claim. Treating those as one step produces instruments that measure what the sensors happen to be good at. Treating them as two, in order, produces instruments that measure what matters, and an honest record of what still isn’t covered.

Koneksa Health brings this same sequencing to sponsor programs: establishing what is clinically and personally meaningful in a given indication first, then designing the digital measurement strategy that can credibly capture it.

Reference: Pearlmutter A, Ellis R, Peterschmitt MJ. Clinician and patient perspectives on meaningful Parkinson’s disease impacts for digital assessment. Front Neurol. 2026;17:1677680. doi:10.3389/fneur.2026.1677680.

Read the full peer-reviewed publication

Click here to access the full resource.

Related Resources

Giving Voice to Early Parkinson’s DetectionCase Study

Giving Voice to Early Parkinson’s Detection

koneksa_podcast_cover_image
Human-in-the-Loop: Why Experts Matter More in the Age of AIPodcast

Human-in-the-Loop: Why Experts Matter More in the Age of AI