- Published
- Updated
Replacing a device during an investigation is more than a logistics transaction. A new unit, different setting or interruption may alter measurement continuity, participant experience and the interpretation of performance. A replacement count alone does not explain these consequences.
Define the assessment
Link the reason, decision maker, old and new units, and their periods of use. Distinguish failure, planned maintenance, participant fit and administrative stock replacement. Technical and statistical reviewers should assess whether a replacement affects data usability rather than assuming it automatically invalidates a participant record.
Worked example
In a hypothetical wearable study, a sensor is replaced because of skin fit and the replacement is worn in a different position. The team records the first valid measurement, interruption and placement difference, not merely the delivery date. It preserves earlier observations and evaluates whether the two measurement periods remain comparable.
Practical checks
- Identify the final valid measurement before replacement and the first afterwards with timestamps.
- Confirm that replacement software, settings and accessories remain within the investigated configuration.
- Link participant instruction, interruptions and safety assessment to the replacement record.
Decision and limitations
The record should explain why replacement occurred and how it could affect the result. Justify analytical handling using the prespecified approach and the actual circumstances. Counting every replacement as technical failure can mislead, but treating every replacement as inconsequential is equally unhelpful.
Frequently asked questions
- Should data from the replaced unit be deleted?
- Not automatically. Preserve the record and assess validity and analytical use through documented reasoning.
- Can planned replacement need safety review?
- Yes. Being planned does not exclude participant impact; the circumstances and any findings still need appropriate assessment.