Answers are grounded in this paper and cite their sources.
Method detection limits estimated from specialized data studies are frequently used as a basis for left censoring reported data. Estimated method detection limits are frequently impacted by the failure to fully mimic all measurement system characteristics represented in process quality data collection. This can result in use of method detection limit estimates in the censoring process that are likely to be biased on the small side. A large set of product quality data analytically provides the substantial combined impact of detection limit censoring in conjunction with common precision and accuracy issues found in classical detection limit estimation processes. A methodology for estimation of practical method detection limits directly from uncensored quality data rather than through a formal detection limit study is provided. The utility and improved accuracy and stability of such practical detection limit estimates versus classically derived detection limits is demonstrated. Guidance is provided with respect to how to approach detection limit censoring whether one chooses to apply it or not to their reported data.
No takes yet. Share an insight, caveat, or question.
Bzik et al. (2024) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: