OMB Meeting Book - January 8, 2015

84

Alternative Approaches to the Traditional Collaborative Study

the likeliness that a validated method will provide equivalent results in the hands of multiple independently operat- ing users at different laboratories. It is sometimes assumed that validated methods can be implemented “straight off the shelf” and achieve the published performance data straight away by whomever uses the method. In truth, the analytical performance of any given method, validated or not, is not known until the method has been verified on- site where the method will be used with the existing equipment and analysts. It is reasonable to expect that a vali- dated method with lower RSD (R) should perform better than a method with an unknown RSD (R) . A collaborative study, although it involves multiple laborato- ries and many factors, does not include all potential sources of variation. So a laboratory must verify a new method to ensure that there are no factors in its laboratory or with its samples that negatively impact the behavior of the new method. On-site verification became a requirement for laboratory accredita- tion after the adoption of ISO 25, a precursor to ISO 17025, in 1999. Today, all accredited laboratories have adopted the practice of on-site verification. AOAC maintains a method verification guideline on its website that describes how to meet the method verification requirements of ISO 17025 (20). With the prevalence of on-site veri- fication of analytical methods, one must wonder if the role of the collaborative study is still as relevant as it once was 30 years ago. Perhaps it is time for another paradigm shift that embraces measurement uncertainty and the on- site verification process. always practical. There are several alternative procedures that might be used to estimate reproducibility that include use of proficiency test- ing data, intermediate reproducibility, and measurement uncertainty. PT data has been found to be equiv- alent to collaborative study data, and a Conclusions and Recommendations Collaborative studies are not

properly designed PT program could be used to determine reproducibility with- out interfering with the principles of PT. Measurement uncertainty is a widely used convention to describe the pos- sible range of results represented by an analytical result. All accredited labora- tories are required to determine and, where applicable, report measurement uncertainty. Measurement uncertainty can be considered equivalent in concept to RSD (R) . ISO has provided guidance for using single-laboratory data to deter- mine the measurement uncertainty of a method. Weitzel and others have dem- onstrated that single-laboratory data can be used to determine measurement uncertainty with proper planning. On-site verification is a common practice for accredited laboratories. It is widely understood that a method cannot be used “out of the box” based on its RSD (R) , but must be verified. The practice of on-site verification reduces the reli- ance on reproducibility results [RSD (R) ].

(7) Weitzel, M.L.J., & Johnson, W.M. (2012) “Using Target Measurement Uncertainty to Determine Fitness-for-Purpose,” Accred. Qual. Assur. doi: 10.1007/s00769-012-0899-x (8) International Vocabulary of Metrology– Basic and General Concepts and Associated Terms (VIM) (2008) JCGM 200:2008, Joint Committee for Guides in Metrology, www. bipm.org (9) Ellison, S.L.R., Thompson, M., Westwood, D., & Wood, R. (2010) “The Role of Proficiency Testing in Method Validation,” Accred. Qual. Assur. 15, 73-79 (10) Thompson, M., & Lowthian, P.J. (1995) Analysts 120, 271-272 (11) Guidelines for the Assessment of the Competence of Testing Laboratories Involved in the Import and Export Control of Food (CAC/GL 27-1997) (12) Ellison, S.L.R., Rösslein, M., & Williams, A. (Eds) (2000) Eurachem/CITAC Guide: Quantifying Uncertainty in Analytical Measurement, 2nd Ed., ISBN 0 948926 15 5, www.eurachem.org (13) Measurement Uncertainty Revisited: Alternative Approaches to Uncertainty Evaluation (2007) Eurolab 2007/1, www. eurolab.org (14) Guidance for the Use of Repeatability, Reproducibility and Trueness Estimates in Measurement Uncertainty Estimation (2010) ISO 21748:2010, International Organization for Standardization (ISO), Geneva, Switzerland (15) Horwitz, W. (1998) “Uncertainty—A Chemist’s View,” J. AOAC Int. 81, 785-794 (16) Federal Agency for the Safety of the Food Chain (Belgium) (November 2008) Estimating Measurement Uncertainty in Quantitative Chemistry Analysis, http://www. favv.be/laboratories/approvedlaboratories/ officecircular/_documents/03-11-2008-proce- dureENLAB-P-508-Measurement-uncertainty- v.01_en.pdf (17) Weitzel, M.L.J. (2012) “The Estimation and Use of Measurement Uncertainty for a Drug Substance Test Procedure Validated According to USP,” Accred. Qual. Assur. 17, 139-146 (18) Weitzel, M.L.J. (January 10, 2012) personal communication (19) Barwick, V., & Ellison, S. (2000) “The Evaluation of Measurement Uncertainty from Method Validation Studies,” Accred. Qual. Assur. 5, 47-53 (20) How to Meet ISO 17025 Requirements for Method Verification (2007) AOAC INTERNATIONAL, http://www.aoac. org/iMIS15_Prod/AOAC/PUBS/GAR/ AOAC_Member/PUBSCF/GARCF/GUIDE_A. aspx?hkey=965b2306-1083-404c-b41a- bf159216a610 ■

—Scott Coates and Deborah McKenzie Contributing Writers scoates@aoac.org dmckenzie@aoac.org —M.L.J. Weitzel Contributing Writer Independent Consultant mljweitzel@msn.com

References (1) AOAC INTERNATIONAL, Guidelines for Collaborative Study Procedure to Validate Characteristics of a Method of Analysis, J. Assoc. Off. Anal. Chem. 72, 694–704(1989) (2) Pure and Applied Chemistry 67 (2), 331-333(1995) (3) J. AOAC Int. 78 (5), 143A–160A(1995) (4) AOAC INTERNATIONAL (July 2011) “Alternative Pathway to Official First Action Method Status Requirements” (5) AOAC INTERNATIONAL (May/June 2011) “AOAC Implements Alternative Pathway to First Action Status,” Inside Laboratory Management, Vol. 15, No. 3, pp 22-24 (6) Youden, W., & Steiner, E. (1975) Statistical Manual of the AOAC, AOAC INTERNATIONAL (reprinted 1987)

24

L A B O R A T O R Y M A N A G E M E N T ■

© A O A C I N T E R N A T I O N A L ■

N O V E M B E R / D E C E M B E R 2 0 1 4