Reagents are typically produced in batches. It is common for there to be some differences between various batches. Manufacturers try to keep these differences as small as possible, but patient results may be affected as a result.
Consequences of lot-to-lot variation:
(1) need to change the normal reference range, which can be a challenge
(2) difficulty to simply track test values over time, since observed changes may be due to the increased test variability
Parameters:
(1) within subject biological variation as percent
(2) number of results that a single patient may have within the reagent lot period = MIN(number, 6), as specified on page 685
(3) measurement uncertainty of calibrator in percent
(4) observed between lot variation in percent
(5) observed pooled average for within-lot variation in percent
Allowable Measurement Uncertainty
X =
= SQRT(((0.5 * (within subject biological variation)^2)) - ((calibrator uncertainty)^2)) / SQRT((number) + 1)
The allowable between lot measurement uncertainty should be <= X
Lot Acceptability
target between lot variation <= (observed pooled average for within lot variation) / (SQRT (number)
If the observed between lot variation is less than or equal to the target between lot variation, then it is acceptable. Otherwise, it is not acceptable.
