College of American Pathologists
Printable Version

  PT approach for nonregulated tests sets
  higher bar






July 2007
Feature Story

Anne Paxton

Ever since the launch of CLIA ’88, differing requirements of the laboratory accrediting organizations have sometimes made federal standards for ensuring test quality confusing.

Some CLIA requirements are clear-cut: Proficiency testing for the analytes regulated by the Centers for Medicare and Medicaid Services (including tests like cardiac markers, blood alcohol, or T3 and T4) is universally required and means five testing challenges, three times per year, must be electronically transmitted to the CMS.

But the “nonregulated” analytes—including tests like hemoglobin A1c or PSA—occupy more of a gray area. Since they’re not waived, they’re subject to CLIA’s external quality control requirement twice a year. Laboratories can meet that requirement by using a proficiency testing program like the CAP’s Surveys, and the College recommends this. But if the laboratory is not CAP-accredited, it technically does not have to use a PT program to satisfy the requirement.

“There are certain analytes where proficiency testing is mandated or regulated by CLIA, but others are not,” points out Leo Serrano, FACHE, CLSup(NCA), director of laboratory services at Avera McKennan Hospital, Sioux Falls, SD (owned by Avera Health), which performs about 2.5 million tests a year. So each approved laboratory accrediting group has had to choose whether to require that nonregulated analytes be subject to proficiency testing.

For some, including the Joint Commission on Accreditation of Healthcare Organizations, the answer is no, in part based on its financial impact on the laboratory. But the CAP maintains that if a proficiency test is available for a test on the laboratory’s menu, whether it is a regulated analyte or not, the proficiency test must be employed. “The College wants to know that you’re doing proficiency testing and comparing yourself to someone else for every analyte where one is available,” Serrano says.

“We have set up with the College all the proficiency tests that it has available, either for regulated or nonregulated analytes,” says Maria I. Mendez, MA, MT(ASCP), formerly the laboratory administrator at the State University of New York Downstate Medical Center, University Hospital of Brooklyn, and now serving as hospital administrator in charge of the laboratory, radiology, and other areas. Mendez says often one specimen can be tested for multiple analytes using the CAP Surveys.

For any other analyte outside the 1,000-plus that CAP offers, “we have to figure out how to validate the test,” she says. For example, for the CH50 test, her hospital performs the test, then sends the specimen to another New York hospital that uses the same method, then compares the results. “Or, we’ll split a specimen here and do it two times and see if we can reproduce the result. Obviously it’s a burden to the laboratory, and it’s a lot more efficient and the data is more beneficial if you subscribe to CAP.”

With the CAP Surveys’ statistical data, “they’re comparing or benchmarking us with our peers in many other institutions, so you know where you fall and how you compare. But obviously another thing,” Mendez says, “is we believe all these materials we are testing are safe; there’s no risk to employees of acquiring any infectious diseases like hepatitis or HIV. We get the material, it’s already tested, we analyze it like any other patient specimen, and we find out how we fit with the median of all the other institutions.”

SUNY Downstate Medical Center’s CAP-accredited laboratory has to meet both New York state and CAP regulatory requirements, and subscriptions to CAP Surveys are critical to making that happen. “The Joint Commission would love to have us change our laboratory accreditation from CAP to them, but we believe the CAP represents laboratories and the pathology profession and understands laboratories better than the Joint Commission,” she says.

The CAP Surveys keep the hospital laboratory personnel on their toes, Mendez adds. “If you get the proficiency test back and it looks like your report is outside the acceptable range, you go back to that day, check the quality control, check the linearity, look at the person who did it and how they handled that specimen—all the preanalytical, analytical, and postanalytical steps. So it does help to look at the whole process and improve quality; we’ve caught transcription errors where the results were missing a decimal point.”

Since enrolling in a CAP Survey might cost $500 or so, some laboratories have tried to save money by doing PT for nonregulated analytes themselves, Serrano notes. But in his view, that’s a bad bargain. His laboratory splits samples only where no survey is available. “There’s a false economy in just splitting samples and swapping specimens back and forth. I’m not a proponent of it. I’d much rather have a survey.”

Avera McKennan also periodically does blind internals. “But that’s different from proficiency testing; all that’s doing is showing you’re repeating the same answer. You’re not necessarily comparing the answer to a broad spectrum of the laboratory community. You won’t recognize or identify any specific shortcomings that would occur within your testing milieu.” He prefers a comparison to several external facilities “to make sure there is not something unique to your facility that is impeding the accuracy of your testing.”

The real problem, in fact, is the nonregulated analytes for which no proficiency test is available, he says. From his experience, any proficiency surveys are a godsend. “It’s been quite a challenge dealing with the newer molecular proteomic and genomic assays, and fortunately CAP has stepped up for some of them. But further complicating the issue is the number of homebrew or local laboratory-developed tests using ASRs [analyte-specific reagents].”

For example, the CYP450 enzyme genotyping test for 2D6, 2CP, and 2C19 alleles is used for monitoring many psychotherapeutic medications, because patients are predisposed to metabolizing the drugs. “Whether you’re a fast or slow or incomplete metabolizer has an enormous bearing on how you react, and on the dose,” Serrano explains. “But to conduct external quality control, we have to find other laboratories that are also doing these tests.”

“It’s been a problem for a while, but it’s become more acute as new tests come on board, and as you’re going through your accrediting process, be it with CLIA or CAP or the Joint Commission or COLA, you may not necessarily have the surveys.”

(The CAP’s Pharmacogenetics Survey includes genotyping for CYP450, among other genotypes, and became available this year.)

In years to come, the importance of using proficiency testing will only grow, in Serrano’s view.

“As you develop more tests, and more virology procedures, and as you expand the scope and breadth of technology available today, you will find more issues with proficiency testing.”

“The cost of quality is not cheap in terms of expense,” he says, in support of the CAP’s required proficiency testing for nonregulated analytes. “But the cost of poor quality is enormous if you’re turning out wrong answers. We would rather be paying for a survey that is properly evaluated and peer reviewed and compared than doing our own thing and having a significantly smaller number of laboratories to compare to.”

Anne Paxton is a writer in Seattle.