Checklist clarifications: To each its own—guidelines for point-of-care QC
Since you asked...answers on POC
Quality control practices that would not be considered ideal in traditional laboratory systems are sometimes appropriate in point-of-care testing, says Stephen J. Sarewitz, MD, checklist commissioner for the CAP Commission on Laboratory Accreditation.
At a July 14 teleconference, Dr. Sarewitz answered three important questions about QC in POC testing that laboratories ask about the CAP Laboratory Accreditation Program checklist.
Sponsored by the CAP and the American Association for Clinical Chemistry, the teleconference focused on POC testing and the Interpretive Guidelines released last February by the Centers for Medicare and Medicaid Services.
Dr. Sarewitz first addressed how tolerance limits are defined for control procedures. A note to checklist question POC.06100 phase II says that for tests with numeric results, recovery ranges applied by manufacturers of assay controls must not be substituted for QC limits determined by the laboratory on its own equipment.
"The question we get in POC testing is, May a laboratory initially or for a period of time use the manufacturer’s ranges, then adjust these if necessary after that laboratory has accumulated a sufficient number of data points?" His answer: Using manufacturers’ ranges is not considered a best practice in traditional laboratory systems, but it is considered acceptable for unit-use devices such as would be used in POC testing.
Most quantitative tests are traditionally monitored with two levels of liquid control material at least each day that patient testing is performed. The daily use of two levels of instrument or electronic controls, or both, as the only QC system is acceptable only for unmodified test systems cleared by the FDA and classified by CLIA as waived or moderately complex.
The concept of external QC has limited applicability to unit-use testing, explained Dr. Sarewitz, who is a staff pathologist at Valley Medical Center, Renton, Wash. "The reason, in effect, is that in unit-use testing, what you’re doing is discarding the analyzer—the cartridge or card—after each test, so you cannot really do meaningful external QC on the analytical component of the test system."
"What you can do is something we call ’acceptance sampling.’ When you do external QC, what you’re really doing is running controls on a sample of a given lot or shipment of cards. If the results of such sampling are acceptable, then you infer that the entire lot or shipment of cartridges is functioning properly." For certain unit-use systems, he noted, the external QC can also test the analyst’s technique.
According to the CAP inspection checklist, if instrument or electronic controls, or both, are being used as the only daily QC system, validation of the system must be documented. The checklist does not prescribe how to do the validation studies, he said. "The reason for this is that, as far as I know, there is no data to support a specific validation protocol that can be applied to all types of unit-use devices in all settings." Thus, each laboratory must develop its own valid protocol based on its setting, its experience, and the particular device it is using. "In many cases," Dr. Sarewitz said, "this validation is really a specific application of the concept of acceptance sampling."
Another phase II requirement (POC.08700) says that if laboratories test more than one instrument for a given analyte, then instruments should be checked against each other at least twice a year for correlation of patient results.
"The question we’re asked is, If you have a POC program with over 100 glucometers, for instance, is it necessary to run patient samples on all these instruments and cross-check all these instruments? That is very difficult to accomplish."
A new note to the checklist clarifies that single-use devices such as POC glucose devices are a special case in which large numbers of devices may be in use, and in this case alternative approaches are acceptable. "One approach a laboratory can take is to document for a lot of reagents or cartridges agreement between patient results for a central laboratory method and representative POC devices."
"The laboratory can then collect QC results from those representative devices and compare them to QC results done on the other POC devices," Dr. Sarewitz said. "If the QC results from the representative group are consistent with the other devices, one can infer that the other devices are appropriately correlated."
All lots of reagents or cartridges that are in use should be included in the group of representative devices that are directly compared to main laboratory instruments, he added. By rotating devices used for each biannual comparison, eventually the laboratory will directly compare all POC devices with the main laboratory instrument.
Another common concern relates to the analytic measurement range, or AMR, the range of analyte over which calibration is valid for the instrument. The AMR must be established for every device when it is put in service, but must also be revalidated every six months. "Again, if a program has 100 POC devices, is it necessary to do a semiannual AMR validation on every device?"
The checklist note on AMR validation was modified recently to provide an exception if there are numerous unit-use devices in a POC program, Dr. Sarewitz said. In that case, the AMR must be validated for every device initially, but then the semi annual AMR validation can be performed on a representative subset of devices.
CAP president Mary Kass, MD, speaking at the teleconference in July, applauded the increasing interest in POC testing. "Not that long ago, it was very difficult to convince pathologists and medical technologists that this is where laboratory testing is going," she said. That 200 sites decided to participate in the teleconference demonstrates "we’ve come a long way in the last five to seven years of POC testing."
Anne Paxton is a writer in Seattle.