Review trial wearable data weekly
- The problem
- Study teams check wearable exports, diaries and labs by hand, one participant at a time, so missed wear time and unusual changes surface late.
- What the demo shows
- A weekly review of real daily wearable summaries from public data: wear-time gaps, changes from a participant's own baseline and out-of-threshold values, each with the sample rule behind it, waiting for a person to decide.
- Why it matters
- The first pass of data review is done for the team, and a person still makes every call, with each decision logged.
Sample study review
Loading public data...
Adherence by site
Share of participant-days at or above the wear threshold set on the Signals tab. Days outside a participant's heart-rate export (at the start or end of the period) are left out rather than counted as low wear.
Participants
Select a participant to open their signals.
Daily signals
Sample review rules
Change these illustrative thresholds and the flags in the review queue update.
Baseline is each participant's own first 14 days. A change is flagged when the 7-day average moves further from it than the setting above. Flags are for review only.
Review queue
Nothing here with these filters.
Select an item to see its evidence and decide.
Audit trail
Every decision in this session, newest first. Kept in this page only and cleared when you reload.
| Time | Role | Action | Item | Reason |
|---|
Weekly digest
Built from the decisions so far. Items not yet decided are marked as drafts.
Concept demo · public wearable signals (Fitbit dataset, CC BY 4.0) in a fictional study. Source: Crowd-sourced Fitbit datasets 03.12.2016-05.12.2016 (Furberg R, Brinton J, Keating M, Ortiz A), Zenodo, 2016, CC BY 4.0, doi.org/10.5281/zenodo.53894.
Where this goes next
On a real study, the same weekly loop runs on the team's own exports with thresholds taken from the protocol, and a person still decides on every item.