Lot-to-lot variation and verification

Lot-to-lot variation affecting calibrators and reagents is a frequent challenge that limits the laboratory’s ability to produce consistent results over time. This variation is not without clinical consequence and there are several well-documented examples of adverse clinical outcomes. It is important that laboratories have procedures in place for quantification of this inaccuracy, and for determining whether the amount of variation is acceptable for the release of patient results. Various approaches have been taken to the assessment of new lots, including the evaluation protocol published by the Clinical and Laboratory Standards Institute (CLSI). Internal quality control and external quality assurance material is often not commutable, and so the use of native patient samples is preferred. Published evaluation protocols differ significantly in ease of use and statistical rigour, and some may be underpowered to detect a clinically meaningful change between lots. Furthermore, current protocols (including the CLSI protocol) will not detect cumulative shifts between reagent lots. This shortcoming may at least partly be addressed by laboratories adopting moving patient averages or similar quality procedures. Collaboration and data-sharing between laboratories and manufacturers also has an important role to play in the detection of lot-to-lot variation. While the laboratory may take steps to evaluate and detect variation, the ideal is to reduce variation between lots at the point of manufacture. Using appropriate acceptance criteria based on medical need or biological variation requirements instead of some arbitrary percentage may go some steps toward achieving this.