- Published
- Updated
The same commercial device name does not guarantee the same configuration at every site. Software versions, sensors, accessories and user settings may change measurement or operation. Make the investigated combination understandable from the study records.
Define the assessment
A configuration record should show components alongside permitted setting ranges. A general manufacturer compatibility list may cover more combinations than the investigation supports. Assess an update through its changed functionality and potential impact, rather than treating a version identifier as a complete explanation of the change.
Worked example
For example, two sites use different image-enhancement settings on an imaging device, each believing it has the default. The team compares recorded examples, links settings to source images and assesses their effect on evaluation. It retains original files and processing history instead of silently reprocessing all historical images.
Practical checks
- Capture the actual hardware, software and important settings during controlled site installation.
- Compare permitted configurations with combinations actually used at sites throughout conduct.
- Record post-change testing, training and data-comparability decisions together rather than in disconnected logs.
Decision and limitations
The trace is useful when the combination generating participant data can be identified. Not every small difference is clinically important; derive significance from effects on measurement and risk. Explain the limits of conclusions where a configuration remains uncertain or insufficiently assessed.
Frequently asked questions
- Is an identical model number sufficient?
- No. Important software, accessory or setting differences can occur without a change in the model number.
- Does a configuration difference invalidate data?
- Not necessarily. Assess with evidence how the difference affects measurement, interpretation and the intended analysis.