Medical Device Daily Washington Editor
GAITHERSBURG, Maryland — The trick for the modern government agency is knowing when to insert itself in the process and when to get out of the way. This was the undertone in last Friday's meeting of the FDA's hematology and pathology devices advisory committee, which examined whether analytical equipment designed to render a complete blood count (CBC) and a differential blood cell count (CBC/diff) should be permitted in the absence of a clinical lab professional.
According to FDA's meeting documents, no waiver for this type of equipment has been granted up to now because no manufacturer has made a cell counter "that is capable of successfully controlling the many sources of analytical error that can result from the lack of oversight by a laboratory professional," even when a physician is present.
Such tests are potent detectors of bacterial and viral infections as well as of anemia and some clotting disorders, but the uncertainty over competence in the modern medical practice led the panel to conclude that this class of analytical devices should remain in the hands of lab professionals.
Paul Rust, VP for point-of-care testing for Quest Diagnostics (Madison, New Jersey), said during the open public hearing that "giving physicians a choice of an overnight test and a point-of-care test" offers "advantages for patients who might find it difficult to get to the doctor for repeated appointments.
"We don't think point-of-care tests should replace central testing all the time," Rust said, but noted that there are circumstances in which a rapid turnaround is desirable. He said one of the firms' subsidiaries has an analyzer that "produces a white blood cell count that is much like when a clinician looks under a microscope," but turnaround is much faster than sending a sample to a lab.
Rust also stated that one physician using the system reduced antibiotic usage from 46% of pediatric patients to 6%, but this doctor followed up to ensure the use was appropriate.
"Is it possible for a pediatrician to operate a CLIA-waived system? We believe it is," he said, adding that the presence of such analytical equipment in a doctor's office can relieve parents of the dilemma of trying to reduce the needless use of antibiotics on the one hand, and the requirement on the parts of some school systems that returning students be inoculated on the other.
Josephine Bautista, associate director of hematology devices at the Office of In-Vitro Diagnostics (OIVD), said "this is important because this is a first-of-a-kind device" for waiver, noting that most CBC counters are for professional use because of the "level of experience needed to interpret" outcomes.
"We have issues such as how can errors be mitigated" in the physician office setting, Bautista added, posing the question of whether "an untrained user [can] interpret and analyze these results?"
Robert Becker, chief medical officer at the OIVD, told the panel that one point of concern for FDA is that "some instruments provide only aggregate" measurements, lumping together such items as neutrophils and eosinophils.
He noted that FDA's guidance on waived tests requires that the assay be accurate in the waived setting and that no reasonable risk of harm to the patient accrues in the event of incorrect performance.
Becker said the challenges for automated cell counters include invalid outcomes due to clots in the sample and short draws. "Hematology results can be affected by small clots that do not affect other tests," he said, but atypically shaped cells can also render invalid outcomes.
"Cell counts alone are not the whole story," Becker observed, stating that raw cell counts will not tell a clinician whether the patient has malaria or some types of severe bacterial infections.
Judy Yost, director of the division of labs at the Centers for Medicare & Medicaid Services, said, "CMS strongly supports point-of-care testing" due to improved access. However, she said that CMS is concerned about the lack of experience often found in doctors' offices.
"The CBC is currently considered a moderate complexity device" unless abnormal cells are detected, at which point it becomes a high-complexity situation, Yost said. She stated that in 2006, about 31% of the labs CMS visited "are not following the manufacturers' instructions" and that high turnover among medical practice staff implies a recurring competence question.
During the panel discussion, panelist John Koepke, professor of pathology at Duke University Medical Center (Durham, North Carolina), said the typical doctor's office has a turnover in staff that "is just amazing.
"Usually, you're trained on the spot," he said of physician office staffers who might be handling CBC counters, adding that manufacturers have to offer ongoing training and competency check-ups before he could recommend a waiver.
The first of FDA's questions was whether CBC and CBC differential testing meet the criteria of "simple" and "insignificant risk of erroneous results."
Panelist Brian Bull, MD, chairman of the department of pathology at Loma Linda University Medical Center (Loma Linda, California), said, "I don't think there's any hope of such a test being simple ... and there is a very significant risk of erroneous results." He said the turnover in physician practices argues against the prospect of the necessary expertise unless routine training is required.
Linda Sandhaus, MD, an associate professor of pathology at Case Western Reserve University School of Medicine (Cleveland), said the question of sample clotting made an "aye" vote tough to go along with, because "its difficult to see how this source of error can be eliminated [when] it has not been eliminated in the lab setting."
However, Helen Wang, MD, director of cytopathology at Beth Israel Deaconess Medical Center (Boston), said the question of whether the samples will settle and hence lead to invalid results is not as conspicuous as other panelists seemed to think. "Believe it or not, the machine does rotate the specimen. I do agree there are a number of pre-analytical issues that can render spurious results, but it's pretty automated," she said.
Wang said if the equipment "just locks out and does not give any results [for outcomes that are not explained by the standard results], it may be worth considering," especially if a second sample goes to a full lab.
However, Koepke countered with the comment that "the current analyzers in the market right now vary," in that not all of them mix the specimens.
Panel chairwoman Dorothy Adcock, MD, medical and lab director at Esoterix Laboratory Services (Englewood, Colorado), said "the panel generally believes that CBC testing is not simple" and can lead to erroneous results, "but this could change" if manufacturers can get past the problem with settled samples and other sample-handling problems. "Then the panel generally believes that waived testing may be applicable to such instrumentation," she summarized.
On the question of how a CBC counter should deal with out-of-specification outcomes triggered by abnormal cells, Sandhause said "current hematology systems do not identify specific intereferences, only what sort of intereferences might be indicated."
Wang countered by noting that CBC counts are "a very common test and I doubt whether every such test is examined" by a lab technologist. She said "this test was done about 88 million times last year and most of them were not scrutinized by a technologist." The situation can be handled by another sample from the patient for further examination.
Adcock said the panel "generally believes that there are many issues that would face an instrument" and "would have to have very good failsafe mechanisms" to account for these.