METHOD CORRELATION OR MADNESS? By Suzanne H. Butch, MA, MT(ASCP)SBB First, let me say that quality control is important. As laboratorians, we have a responsibility to follow federal regulations. Furthermore, we have a responsibility to bring to the attention of our regulators when there are illogical and time consuming quality control requirements that fail both the purpose and spirit of quality management. One such example has recently come to my attention. In the newest (6/15/09) CAP Transfusion Medicine Survey is question TRM.31450 that asks “If the laboratory uses more than one instrument/method to test for a given analyte, are the instruments/methods checked against each other at least twice a year for correlation of results?†The question and the explanatory note below is word for word the Chemistry/Toxicology question CHM.13800: “NOTE: This requirement applies to tests performed on the same or different instrument makes/models or by different methods. This comparison must include all nonwaived instruments/methods. The laboratory director must establish a protocol for this check. Quality control data may be used for this comparison for tests performed on the same instrument platform, with control materials of the same manufacturer and lot number. Otherwise, the use of fresh human samples (whole blood, serum, plasma, urine, etc.) rather than stabilized commercial controls, is preferred to avoid potential matrix effects. In cases when availability or pre-analytical stability of patient/client specimens is a limiting factor, alternative protocols based on QC or reference materials may be necessary but the materials used should be validated (when applicable) to have the same response as fresh human samples for the instruments/methods involved. This checklist requirement applies only to instruments/methods accredited under a single CAP number.†My interpretation of the above is that every six months blood bank laboratories must now document correlation between all methods used to perform a Type, Screen, Crossmatch and antibody identification whether by tube, manual or automated microtiter plate, manual or automated Gel and by different phases. We must document that we get the same answer, regardless of method. The rationale behind this CAP question is that correlation studies are a required by CLIA regulation. This appears to be an unintended consequence of the most recent revision of the CLIA quality control regulations where individual laboratory requirements were eliminated in favor of a single set of requirements. This requirement makes sense when the results being reported vary from day to day and instrument to instrument. However, blood types do not change without cause. They are not variable from day to day. In addition, the ABO type is “controlled†by the use of a forward and reverse grouping. Tube Rh typing result vary by the clone(s), enhancement media, incubation time, cell suspension and other patient variables. It is well known that the various methods used for antibody detection and identification have differences in sensitivity and specificity. In addition, some antibodies are only recognized under very specific circumstances. No two antibody detection/identification methods get the exact same results. In fact, we employ this variation in methods to obtain different results when we use enzymes, PEG, LISS, etc. to help us problem solve when a patient has multiple antibodies, non-specific reactivity, and warm and cold auto antibodies. The most important “control†we do in the transfusion service is the history check. We do this for every patient. If we find a discrepancy, we investigate. Doing a method comparison every six months will not improve patient care. Applying what is a rational requirement when doing biochemical tests to the serologic results produced by immunohematology testing is not appropriate. If the method comparison was easy to perform, one might decide to just comply. In this case, however, meaningful testing is cumbersome, difficult to structure and execution is expensive to execute. Significant time and resources would need to be devoted to this task. Testing enough samples to provide a valid comparison of methods is time consuming and illogical. The references given for this new requirement are based on biochemical and hematological studies: 1) Department of Health and Human Services, Centers for Medicare and Medicaid Services. Medicare, Medicaid and CLIA programs; CLIA fee collection; correction and final rule. Fed Register. 2003(Jan 24):5236 [42CFR493.1281(a)] 2) Podczasy JJ, et al. Clinical evaluation of the Accu-Chek Advantage blood glucose monitoring system. Lab Med. 1997;28:462-466 3) Ross JW, et al. The accuracy of laboratory measurements in clinical chemistry: a study of eleven analytes in the College of American Pathologists Chemistry Survey with fresh frozen serum, definitive methods and reference methods. Arch Pathol Lab Med. 1998;122:587-608 4) Miller WG, Myers GL, Ashwood ER, et al. State of the Art in Trueness and Inter-Laboratory Harmonization for 10 Analytes in General Clinical Chemistry. Arch Pathol Lab Med 2008;132:838- 846 5) Clinical and Laboratory Standards Institute. Verification of comparabililty of patient results within one healthcare system: Approved Guideline. CLSI document C54-A (ISBN 1-56238-671- 9).Clinical and Laboratory Standards Institute, 940 West Valley Road, Suite 1400, Wayne, Pennsylvania 190871898, USA, 2008. For those of you who share my opinion and believe the arguments against method correlation on a 6 month basis are cogent, please write Judy Yost Director, Center for Laboratories US Dept of Health & Human Services Commission on Medicare & Medicaid Baltimore, MD or CLIA staff at (410) 786-3407 or (410) 786-3531. Her email is Judith.yost@cms.hhs.gov . I have heard her speak and she is a proponent of reasonable and effective quality control.