.jpg){kind=link}
- Published
- Updated
Different units for the same test can create misleading differences during review. Treating conversion as arithmetic alone may conceal distinctions in test identity, specimen type or measurement method that matter to interpretation.
Preserve the meaning of data
Link every conversion rule to the original test code, source unit and intended use. Scientific suitability of a factor and correct software implementation are separate questions. Identify equivalence decisions requiring laboratory expertise. Do not silently map an unfamiliar unit to the nearest available option, and retain the source number together with its unit.
Worked example
For example, two laboratories report an analyte in different units. A mistaken mapping for one laboratory produces technically valid numbers at the wrong magnitude. The team compares realistic low and high results against independent calculations and identifies records processed under the incorrect mapping version. Successful file loading would not reveal this error.
Review checks
- Version each factor with its analyte and source-to-target unit pair.
- Test rounding order and lost precision using boundary examples.
- Expose values that cannot be converted as explained exceptions.
Acceptance and boundaries
Acceptance evidence should allow a reviewer to reproduce the calculation from source to destination. Correct arithmetic does not by itself establish that different measurement methods are clinically comparable. If a rule changes, data management and relevant scientific reviewers should decide which historical values need reprocessing. Record that decision rather than overwriting earlier outputs without explanation.
Frequently asked questions
- Is one conversion table sufficient?
- It can be a useful starting point when its scope, approval, version and exceptions are explicit.
- Can the original unit be removed?
- Retain the original number and unit so the measurement remains interpretable and the conversion can be checked independently.