Upper-echelon QA through Accuracy-Based Programs

Create PDF

Anne Paxton

June 2015—HbA1c, creatinine, testosterone, vitamin D, lipids, and maybe albumin. If you know what the common thread is among these analytes, then you may be familiar with the CAP’s Accuracy-Based Programs and their evolution over the past couple of decades.

While a few providers around the world offer accuracy-based surveys, the College’s Accuracy-Based Programs are by far the largest. As the CAP Accuracy-Based Testing Committee looks back on the progress of this set of Surveys and considers new biomarkers to include, it has no trouble showing that the Surveys have been effective in raising laboratories’ awareness of accuracy, improving standardization of tests, and moving manufacturers, correspondingly, to improve their assays.

The current list of Accuracy-Based Programs includes: Accuracy-Based Lipids (ABL), Accuracy-Based Testosterone and Estradiol (ABS), Accuracy-Based Vitamin D (ABVD), Accuracy-Based Urine (ABU), Hemoglobin A1c-3 Challenge (GH2), Hemoglobin A1c-5 Challenge (GH5), Hemoglobin A1c CVL (LN15), and Creatinine Accuracy CVL (LN24).

While proficiency tests serve as a valuable check on the accuracy and reliability of laboratories’ testing, a laboratory’s results are not compared against a gold standard in most such tests. Rather, proficiency testing works on the basis of peer group ratings. “Your individual lab result is compared to the average of all other participants that used the same method,” explains Greg Miller, PhD, a member of the CAP’s Accuracy-Based Testing Committee and professor of pathology and director of clinical chemistry at Virginia Commonwealth University.

“If you have the same value as everybody else using the same method, you can be confident you’re using the method the way it was intended. It’s a good check that the lab has implemented a method and is using it correctly, but it does not tell whether a particular method itself has a bias against a correct value from a reference method or that all methods give the same results for patient samples.”

Dr. Miller

Biases in proficiency testing results can be caused by matrix effects—the effects of all components of the sample other than the analyte of interest—on the measurement of the analyte. Matrix effects can stem from modifications made to a proficiency testing material during its preparation.

“Generally speaking, the matrix-related bias is a property of a method type,” Dr. Miller says. “So if you score an individual against the average of everybody using the same method, you can ignore the matrix-related bias because it affects all users of the same method in the same way.”

While peer group grading has limitations, there’s a practical basis for conducting proficiency testing that way. The reason was that the materials in the proficiency tests were generally stabilized or processed material that could be produced in large quantity, says Accuracy-Based Testing Committee member John H. Eckfeldt, MD, PhD, professor of laboratory medicine and pathology at the University of Minnesota.

“They could test dozens of potential analytes in chemistry. But by adding various things or stabilizing the material, they often introduced what’s called non-commutability. The material no longer behaves like a patient sample in the measurement sample. So biases across peer group results are seen as an artifact of non-commutability. That’s the concern—and it’s left something of a cloud over whether actual patient sample results are accurate or not,” Dr. Eckfeldt says.

“That’s where the Accuracy-Based Programs come in to fill the gap,” Dr. Miller says. “They provide a way to assess the bias between different procedures in a reliable manner.”

Much of the laboratory world takes it for granted—mistakenly—that bias has been systematically rooted out of the testing process. “We find, even today, that a number of international guidelines and clinical textbooks are predicated on the assumption that all methods are producing comparable results,” says Accuracy-Based Testing Committee member Darryl Erik Palmer-Toy, MD, PhD, medical director of Kaiser Permanente Regional Reference Laboratories in North Hollywood, Calif. “So it’s important that we, at the very least, harmonize our test results to make different methods look the same, or better still, produce accurate results.”