Robust estimation of between and within laboratory standard deviation with measurement results below the detection limit
In method validation studies, the question arises how to deal with <LOD results in order to ensure reliable precision estimates. It is proposed to elaborate strategies for dealing with <LOD results on the basis of the Q/Hampel method, a particularly robust statistical method. Two different strategies are presented and assessed: censoring <LOD results and setting them equal to LOD/2. The two strategies are then examined in the light of the reliability of the precision estimates obtained on the basis of a comprehensive simulation study. The first method exhibits better precision, while the second method displays better trueness. However, the two methods are only satisfactory as long as the percentage of <LOD results is less than 25 %. For higher percentages of <LOD results, more sophisticated methods should be applied.
KeywordsLOD <LOD results Method validation Censoring Robust statistical methods Q method Q/Hampel Reproducibility Repeatability
Conflict of interest
The author declares that there are no conflicts of interest.
- Colson B (2015) The Untold History of Algorithm A. http://quodata.de/en/company/algorithm-a.html
- Uhlig S (1997) Robuste Schätzung in Kovarianzmodellen mit Anwendung im Qualitätsmanagement und in Finanzmärkten. Habilitationsschrift, Freie Universität Berlin, Berlin. Available at quodata.de/enGoogle Scholar
- Uhlig S, Colson B (2015) Q/Hampel*: A new robust method for interlaboratory studies with high percentages of “less than” values. (in preparation). http://quodata.de/en/company/algorithm-a.html