Start with the full set of time-stamped readings, the event report and the rule used to classify the event. These records let you trace how the reviewer moved from individual measurements to the reported conclusion.
Build the chronology
Request readings before and after the flagged interval, along with the time zone and any gaps or status flags. Ask for the device identifier, deployment records and any contemporaneous review notes. Keep the time of the first qualifying reading separate from the time the alert was transmitted or reviewed.
Reproduce the classification
Identify which readings satisfied each criterion and how missing or flagged samples were handled. A systematic review of wearable monitors shows that detection sensitivity can vary with the chosen criteria, which makes the rule version part of the evidence to review.
Source
Van Egmond et al., 2020 systematic review of wearable transdermal alcohol monitors
Leave a Reply