Skip to main content
U.S. flag

An official website of the United States government

Comparison of detection limits estimated using single- and multi-concentration spike-based and blank-based procedures

January 29, 2021

Spike- and blank-based procedures were applied to estimate the detection limits (DLs) for example analytes from inorganic and organic methods for water samples to compare with the U.S. Environmental Protection Agency's (EPA) Method Detection Limit (MDL) procedures (revisions 1.11 and 2.0). The multi-concentration spike-based procedures ASTM Within-laboratory Critical Level (DQCALC) and EPA's Lowest Concentration Minimum Reporting Level were compared in one application, with DQCALC further applied to many methods. The blank-based DLs, MDLb99 (99th percentile) or MDLbY (= mean blank concentration + s × t), estimated using large numbers (>100) of blank samples often provide DLs that better approach or achieve the desired ≤1% false positive risk level compared to spike-based DLs. For primarily organic methods that do not provide many uncensored blank results, spike-based DQCALC or MDL rev. 2.0 are needed to simulate the blank distribution and estimate the DL. DQCALC is especially useful for estimating DLs for multi-analyte methods having very different analyte response characteristics. Time series plots of DLs estimated using different procedures reveal that DLs are dependent on the applied procedure, should not be expected to be static over time, and seem best viewed as falling over a range versus being a single value. Use of both blank- and spike-based DL procedures help inform this DL range. Data reporting conventions that censor data at a threshold and report “less than” that threshold concentration as the reporting level have unknown and potentially high false negative risk. The U.S. Geological Survey National Water Quality Laboratory's Laboratory Reporting Level (LRL) convention (applied primarily to organic methods) attempts to simultaneously minimize both the false positive and false negative risk when 

Publication Year 2021
Title Comparison of detection limits estimated using single- and multi-concentration spike-based and blank-based procedures
DOI 10.1016/j.talanta.2021.122139
Authors William T. Foreman, Teresa Lynne Williams, Edward Furlong, Dawn Hemmerle, Sarah Stetson, Virendra K. Jha, Mary C Noriega, Jessica A Decess, Carmen Reed-Parker, Mark W. Sandstrom
Publication Type Article
Publication Subtype Journal Article
Series Title Talanta
Index ID 70219570
Record Source USGS Publications Warehouse
USGS Organization National Water Quality Laboratory; WMA - Laboratory & Analytical Services Division