Lancaster Laboratories Environmental

Method Detection Limits – Current and Future Directions

Users of environmental data frequently ask questions about method detection limits (MDL) and the associated concept known as limit of quantitation (LOQ). Understanding how those values are determined, what they mean and why they might change are key concepts in making analytical method choices and using the results from environmental testing. Many laboratories, including Lancaster Laboratories, use the procedure published by USEPA in 40 CFR Part 136, Appendix B to determine MDLs.

Chemist Lindsey Lafferty, Pesticide Residue Analysis Group, performs method detection limit verification on the group’s GC instruments, which is done annually.

A basic definition of the MDL is the minimum amount of analyte that can reliably be discriminated from a blank. The current USEPA procedure for determining the MDL uses statistics to predict the concentration that can be measured and reported with 99 percent confidence that the analyte concentration is greater than zero and not a false positive. Multiple samples fortified with low levels of analyte at the approximate concentration of the MDL are analyzed, and the standard deviation of those measurements is used to calculate a statistical MDL. However, the statistical MDL can vary due to factors such as the condition of the instrument, analyst skill and spike level chosen. While some laboratories report sample results down to the level of the statistical MDL, Lancaster Laboratories pools data from multiple instruments and analysts, an approach that allows reporting of a consistent MDL for each method and accounts for day-to-day variation in the testing system. For SW-846 organic methods, a confirmation standard is analyzed after each initial calibration to verify that the MDL can be detected under current operating conditions.

Most of the analytical report options offered by Lancaster Laboratories include the LOQ, which may be extrapolated from the MDL and is defined as the concentration above which quantitative results may be obtained with a specified degree of confidence. Values reported below the LOQ have a higher degree of uncertainty associated with them and are usually reported as estimates (J values), when low-level reporting is requested. On an annual basis, Lancaster Laboratories evaluates statistical MDLs on each instrument used for every method employed and type of matrix analyzed and uses that data to determine a single LOQ that accounts for background and instrument variability. A standard at or below the LOQ is analyzed as part of the calibration curve to provide ongoing verification of the values. A variety of situations, including insufficient sample quantity, high concentrations of target compounds and high levels of matrix interference, can make it necessary to increase the MDL and LOQ for specific samples.

Although the MDL protocol published in 40 CFR Part 136 is widely used, it has also been criticized for its limitations. The statistics generated in the current procedure assume a straight-line calibration model and do not directly address the error rate for false negative values. To address these concerns, USEPA established a Federal Advisory Committee on Detection and Quantitation Approaches and Uses in Clean Water Act Programs. After two years of studying the issues and solutions, the committee issued its final report with recommendations for the Agency to consider. The committee did not reach consensus on a method to determine MDLs and LOQs that would provide improved data, but they did recommend a draft procedure that USEPA may approve in the future. The committee’s full report is published at: http://www.epa.gov/waterscience/methods/det/faca/final-report-200712.pdf