Feature Story

Worries surface in wake of waives

February 2001
Anne Paxton

To waive or not to waive? Ever since the Clinical Laboratory Improvement Amendments began taking effect in the early 1990s, that is the question federal officials have asked about hundreds of laboratory tests. Under a CLIA waiver, of course, a test can be performed in physician office laboratories and other settings without meeting CLIA quality control requirements that apply to tests defined as moderate- or high-complexity.

Many physician groups, most notably the 294,000-member American Medical Association, support a longer list of waived tests and want to be sure there is no tightening of the criteria used to determine which tests qualify for waiver. But the College and other laboratory organizations have grown increasingly alarmed about how the waiver process is being conducted.

The number of waived tests has surged, and the climb started even before the Food and Drug Administration took over responsibility for CLIA waivers from the Centers for Disease Control and Prevention in January 2000. "There has been a very, very significant increase in the numbers of tests placed into the waived category over the past two or three years," says CAP president Paul Bachner, MD, a former member of the Clinical Laboratory Improvement Advisory Committee and chairman of the Department of Pathology and Laboratory Medicine at the University of Kentucky Chandler Medical Center, Lexington.

"In part this reflects a number of ’me-too’ tests, but we’re also starting to see different types of tests becoming waived." These decisions have not really been based on any scientific study, he adds. "There’s a sense that the bar has been lowered."

Glen Hortin, MD, chair of the College’s Point-of-Care Testing Committee and chief of clinical chemistry in the Department of Laboratory Medicine at the National Institutes of Health, agrees. "There’s been a rapid explosion in the number of waived tests," Dr. Hortin says. "Potentially, the way things were going, almost any test could evolve into a waived form."

Relaxed rules?

Along with nine other laboratory organizations, the College called attention to this apparently new policy direction in a recent letter to then Health and Human Services secretary Donna Shalala. "We are concerned that the FDA may have changed the criteria for placing tests in the waiver category without reaching an agreement with CDC and [the Health Care Financing Administration] on how to proceed, nor seeking the advice and input from CLIAC, the agency advisory panel on laboratory issues," the co-signers stated.

The quality of waived testing has long been a source of worry for the College, which believes that the whole category may pose a potential hazard to patients. In the CAP Laboratory Accreditation Program, all testing is considered to be of equal value, and the same quality standards are applied whether or not the tests are waived under CLIA.

"The College is not opposed to the concept of waived testing," says Dr. Bachner. "But we strongly believe that any testing that’s done for purposes of patient care should have quality control and proficiency testing performed. That was our position prior to CLIA and continues to be our position."

Some of the certificates of waiver that have been granted recently are more surprising than others, says Dr. Hortin, citing a test for detecting bladder cancer as an example. "But I’m not sure people have identified any particular test as the primary problem. It’s largely a matter of the tests moving out into a relatively unregulated environment, where there’s some evidence that standards of practice have not been appropriate in terms of performing the test."

In an October 2000 letter to the FDA, the College defended the principles that waived tests should be limited to those that do not include complex chemical or biological methods, that the automatic categorization of home-use tests as waived is inappropriate, that the tests must be used in a responsible fashion in whatever setting they happen to be deployed, and that all test procedures should be subject to a documented QC program.

The FDA should hold manufacturers accountable for incorporating QC and PT into waived test device design, and HCFA should identify and address providers’ noncompliance with manufacturers’ instructions, the CAP believes.

But organizations representing the roughly 75,000 POLs that confine their laboratory testing to the waived provider-performed microscopy categories, along with manufacturers of testing devices, have not been shy about expressing an opposing view. "It’s a fairly contentious issue," says Clara Ann Sliva, MT(ASCP), MPA, acting CLIA coordinator in the FDA’s Division of Clinical Laboratory Devices.

Sliva is now at the center of an FDA effort to resolve the controversy and bring predictability to the waiver-granting process by finalizing a long-pending rule on exactly what the criteria mean. Her division, which sponsored a workshop on Aug. 14-15, 2000 to hear all sides’ concerns, may decide to continue to apply the current criteria, finalize a proposed rule published by the CDC in 1995, or re-propose other procedures and criteria for the process.

As a first step, the division on Jan. 23 released a draft guidance document on waived testing (www.fda.gov/ cdrh/ clia). "It’s 2001, and we need to take a fresh look" at the criteria, she says.

De facto standard

Nearly 13 years after CLIA’s passage, the criteria for issuing a certificate of waiver have yet to be spelled out. The CLIA legislation designated for waiver eight tests that were commonly performed in a home setting, such as pregnancy and glucose tests, but the mechanism for waiving additional tests was left unspecified.

Their home-use history led to those tests being "kind of shoehorned in under the definition," because they were considered to be safe and not really being used for diagnosis-just for screening-and were to be followed up by a more definitive test, Dr. Hortin says. Definitions of the crucial terms "simple" and "accurate" in the list of criteria were left for another day.

Nevertheless, between 1994 and Jan. 30, 2001, 733 tests were waived. The majority-about 80 percent-were tests like fecal occult blood, pregnancy, ovulation, and urine dipstick that were waived by regulation, Sliva says.

At first, the CDC, the agency then in charge of granting waivers, relied on the conservative advice of CLIAC that only "near perfect" or foolproof tests should be waived, she explains. Later, in September 1995, the CDC and Health Care Financing Administration jointly issued a proposed rule on waived-test criteria, which has been considered current policy though it hasn’t been finalized.

In 1997, with the FDA Modernization Act, "there was what we called a ’CLIA adjustment,’" Sliva says. The act’s language clarified home-use tests as waived and provided direction with examples describing waived tests.

Since the FDA’s "takeover" of waived testing from the CDC, the first new analyte to be waived was Quidel Corp.’s QuickVue influenza test, followed by ZymeTx Inc.’s ZstatFlu. "But we also have other analytes that have come through the over-the-counter criterion, like follicle-stimulating hormone, and another one for estrone-3-glucuronide," Sliva notes.

Frank R. Rudy, MD, a commissioner of the College’s Laboratory Accreditation Program and president of Pathology Associates of Central Pennsylvania in Harrisburg, served as the CAP’s spokesperson at the August workshop. He believes the FDA is concerned that the waiver criteria have not been applied consistently because they may be too subjective.

But, he adds, "It’s my impression that the FDA hearing was really prompted by industry, which wants to have the waiver process be as easy and quick as possible. There are several manufacturers and marketers who have had tests in the application process for protracted periods of time, and wish to have some final action either yea or nay."

William Hamlin, MD, chair of the College’s Commission on Laboratory Accreditation, says it is unclear why the government has granted more CLIA waivers recently, but he thinks several factors are contributing to the trend. "I think probably the nature of testing is changing. Tests are becoming much simpler and don’t require a lot of equipment. Point-of-care testing is being done in a lot of different places, many in POLs and all sorts of venues all over the country. The manufacturers in general are producing more tests and it’s to their advantage to get the tests waived."

In fairness, Dr. Rudy says, some waived test applications have apparently been approved in a timely fashion while others have lingered in the system for years. "It does appear that CDC’s criteria were not consistently applied." But the College’s concern, he says, is that "a waiver in essence means the tests are not subject to required QC, they’re not subject to peer review, and the laboratories performing these tests are never subject to inspection."

Calculating the risk

At the FDA workshop, Dr. Rudy pointed to several examples of quality problems in waived testing that were discovered through CAP inspections or oversight. One laboratory found while evaluating nurses’ performance that a nurse was placing the dipstick in the urine, then immediately reading the results without following prescribed waiting periods for the different components of the strip, causing false-negative results. "This lapse was only identified because this was a CAP laboratory," he points out. "But how many patients [in that institution] had delays in diagnosis of urinary tract infections, hematuria, or glycosuria?"

In another case, a health care provider was using two different waived test kits for fecal occult blood. The developer reagents for the kits were not interchangeable, yet the reagent from one kit was being used with the fecal occult card from another kit. Furthermore, not being laboratorians, the users did not pay attention to the fact that the positive control on the test card was negative. "One individual thought the reagent used with fecal occult blood was water, so after it was empty he filled it with tap water and used this as his reagent system. Needless to say, all of his results were negative. How many patients had undiagnosed gastrointestinal bleeding, polyps, or colon cancers?"

A glucose meter in use at a neonatal intensive care unit illustrated the need for ongoing quality control, Dr. Rudy continues. "It had been functioning normally, but abruptly failed QC testing, so it was immediately removed and evaluated," he recounts, noting that the meter failed linearity testing miserably. "How many neonates would have been harmed if the problem had gone undetected for even a short period of time? In a waived testing laboratory QC would not have been performed, so the device failure would have gone undetected. These meters can and do fail."

"A lot of people look at the devices and believe they’re so simple they’re failure-proof. In reality, they’re not," he says, noting that more problems have arisen from glucose meters than from any other testing device.

Some manufacturers objected at the workshop that these examples were just anecdotes. "We did come under attack, as I anticipated, from industry saying that regulating by anecdote and exception is inappropriate because in all likelihood it’s not reflective of the general quality of the tests performed in the community," he says. But after Dr. Rudy’s presentation, a HCFA official reviewed recent pilot studies that the Medicare agency performed that confirmed Dr. Rudy’s critique of waived testing quality.

A random HCFA inspection of 80 to 100 laboratories in Ohio and Colorado-mostly POLs-found that 30 percent in one and 50 percent in the other were not appropriately performing tests. "To make sure these problems are not limited to those regions, HCFA said it intends to conduct a random review by inspecting laboratories in each of the other eight regions of the country," Dr. Rudy says.

But even during the workshop, a member of the New York state health department commented that a New York evaluation of 600 laboratories performing waived testing found that more than 50 percent were not performing tests appropriately, usually because of not following manufacturers’ instructions.

The AMA stressed in its workshop presentation that CLIA, in its view, has not significantly improved the quality of laboratory testing, has caused thousands of physicians to abandon office testing, and is a major contributor to the 110,000 pages of regulations that the federal government has imposed on the practice of medicine.

Urging the FDA to retain the current criteria for determining waived tests, the AMA said it opposes any new regulatory burden on POLs. Further, the organization stated, since manufacturers and the physician community have relied on the understanding that CLIA would not be significantly modified, "changing those rules now would be a distinct disservice."

While not commenting specifically on the HCFA and New York studies, the AMA did say that if they are accurate, "We believe that physician education is the solution, not further restrictions on the waiver category." If there is a problem with manufacturer instructions, then HCFA and manufacturers should focus on educating users regarding proper use of equipment, the AMA asserted.

Fifteen medical groups co-signed the AMA statement of opposition to restricting waivers, among them the American Academy of Family Physicians, American College of Cardiology, Medical Group Management Association, American Society of Internal Medicine, and American College of Obstetricians and Gynecologists.

Access versus quality?

That the College does not recognize a waived category in its accreditation program reflects the CAP’s conviction that the same standards should apply for testing across the board, Dr. Hamlin points out. "We require that all tests be quality controlled and have proficiency testing done or an alternative serving equally well. The accreditation program has taken that position since its inception. In that sense, we’re more stringent than CLIA." The College believes no incorrect test result is "risk free," Dr. Rudy explains. Correspondingly, the College contends that any test that can’t harm a patient can’t help a patient.

For the same reason, the College believes that automatically categorizing home-use tests as waived is a mistake because home-use criteria set a lower standard. While the College supports FDA approval of home-use tests for patient use, it questions the use of these tests in multiple sites by a variety of unskilled users with no oversight or quality assurance.

But several medical groups contend that the high quality the CAP champions may come at the price of patient access to testing. According to device industry estimates, fewer than half of the primary care physicians in the United States have CLIA moderate- or high-complexity laboratories in their offices. Indeed, data from HCFA indicate that over the past several years, 5,000 POLs have dropped moderate- and high-complexity testing from their menus-in many cases, presumably to gain exemption from required proficiency testing, inspections, and other regulations.

Would the same pattern hold if waiver criteria became more restrictive? Dr. Rudy believes not. "Their argument is that these POLs could not and would not perform this testing if they had to perform quality control, be subject to peer review, or be inspected. From the FDA’s perspective, this is an access-to-care issue. I do not believe," Dr. Rudy says, "that physicians responsible for the oversight of waived testing would object to the performance of quality controls and proficiency testing if they understood their critical importance and the risks associated with relying on test results in the absence of QC and PT. It’s largely a matter of education."

The AMA argues that CLIA requirements cut patient access to on-the-spot testing. But Dr. Hamlin disagrees. "I’ve been in laboratories all my life, and I don’t think requiring some appropriate control for this testing as well as proficiency testing would in any way affect access or closeness to the patient."

Physician groups and manufacturers have pushed for minimal regulation of office laboratories, "but most laboratory groups have been more conservative," Dr. Hortin adds. "They’re concerned that we’re undergoing a large experiment here, and we really don’t know what the consequences will be as you move tests out of a relatively regulated environment where there are laboratory professionals performing the test, and the usual lab standards and regulations apply, into a setting where you have highly variable personnel, some with limited laboratory training and usually no participation in proficiency testing or other ways of checking up on whether quality is being maintained. There’s a question as to whether you’re going to get the same quality of product."

He is somewhat less concerned about waived testing performed in the hospital setting. Because of Joint Commission or CAP requirements for point-of-care testing, he points out, "some minimal standards usually apply even if the test is classified as waived."

But a significant drawback of the fact that more tests are being waived is that data on quality, which could shed light on the controversy, are harder to obtain. "Moving to an unregulated environment means you have lost the ability to capture data," Dr. Hortin notes. "You’d really like to know what the impact is on test quality, but these laboratories tend to be lost out of our usual data collection mechanisms."

A matter of perception

Dr. Bachner suggests there is merit to some of the AMA’s concerns about waived testing criteria. "For a number of years there was intense unhappiness with CLIA ’88 on the part of organized medicine. Going beyond their basic philosophical feeling that CLIA was an unwarranted federal intervention, both the AMA and several of the major professional medical societies felt from a practical standpoint that it was creating an access problem, and making it harder for physicians to offer laboratory tests in a way to benefit patients."

Primary care groups, however, have not really engaged with the implications of the HCFA studies or with earlier research showing that proficiency testing performance varies according to testing site, Dr. Bachner believes. "They don’t necessarily contest the findings per se. Their position has been that in the physician office the performance of the test is subject to a level of control or knowledge about the patient that doesn’t exist in the laboratory that’s remote from the site."

"When you talk to physicians about testing in POLs, what you often hear is ’We don’t really care whether the result is accurate to so many decimal points. We’re more interested in a reasonable number that we can base the care of the patient on. Our results are close enough for our clinical purposes.’" Physicians often don’t fully appreciate that even professional laboratories do better when they are subject to regulation and oversight. "A laboratory that is inspected by the College in general will have more deficiencies identified on the first cycle than they will on their second cycle," he stresses.

For its part, the public tends not to understand the risks of waived testing as much as that of other types of testing-cytology, for example, Dr. Rudy says. "Cytology received an incredible amount of media scrutiny, so clearly there is a heightened patient perception of the risk of erroneous Pap tests," he points out.

In addition, a missed diagnosis of cervical cancer clearly catches the attention of the media more quickly than an undiagnosed infection. "But is an erroneous Pap result more or less significant than a false-negative fecal occult blood test where a patient’s colon cancer went undiagnosed? Clearly not. So errors in waived testing can be as significant as errors that may occur in non-waived testing."

The reception to the Institute of Medicine study on medical errors last year demonstrated that the public can be made to understand the potential danger, he says. "It’s my feeling that if the public became aware of the error rate and problems in the performance of waived laboratory testing, they would find that risk unacceptable," Dr. Rudy maintains. "It would be difficult for the populace at large to understand why a POL wouldn’t perform quality control, when it’s well recognized that those endeavors increase the reliability of testing and help identify testing errors." The public believes, he notes, that the reliability of a laboratory result is the same regardless of where the test was performed. "This may not be the case," he adds.

Despite the widespread attention to the IOM report, Dr. Hamlin is not sure how well the public understands quality issues. "But from the complaints we see from time to time about laboratory errors or misreports during the course of accreditation, I can tell you the people doing the complaining are very concerned."

"When they go to a doctor’s office laboratory and get something done, patients have a higher level of confidence and expect the answers to be right, and the implications in patient care are basically the same whether the test is done in the POL or sent to a central laboratory," Dr. Hortin says. "CAP’s position essentially is that as long as the test is being used for diagnosis or patient management, there is no kind of test that if you don’t perform it right, it doesn’t matter."

The push for CLIA repeal by physician groups has faded, Dr. Bachner says, with the growth of the waived test list and the addition of different types of testing to the menus, and with regulations on provider-performed microscopy that allow nonphysician staff to perform tests. Those factors "have very significantly taken away some of the steam on the part of physician groups to exempt POLs entirely from CLIA," he says.

Informed sources in Washington, DC, confirm that while organized medicine still supports CLIA’s repeal, leaders privately concede that the law is firmly entrenched and the likelihood of Congress reviving a pre-CLIA era is probably remote. Their priority instead is to find further relief from the law’s requirements.

Down the road

Among several professional organizations, there is still uncertainty about the direction the FDA will take. "At the August workshop, they tentatively said they would slow down and wait until people decided what was going on, but then some people were surprised when in October they approved a new category of point-of-care tests for influenza," Dr. Hortin says.

Meanwhile, the Jan. 23 guidance document, released at CAPTODAY press time, was slated for discussion at the Feb. 7-8 CLIACmeeting, the focus of which was to be the waiver question. "The guidance document was published in record time," Sliva notes, though she predicts it will be 2003 before a new rule on waivers is adopted.

"The rule itself will be very high priority for HHS," she says. The agency "is committed to obtaining appropriate input and moving forward" quickly.

Dr. Rudy says the College recognizes and supports the advances in device technology and the need for point-of-care testing in a physician’s office, at the hospital bedside, or in the patient’s home. "But we do not believe the reliability of the test result should vary by site of performance. The argument should not be quality versus access, for in health care quality should not be knowingly compromised or sacrificed. Rather, we should keep the focus on access to quality care."

Anne Paxton is a freelance writer in Seattle. For a complete copy of the College’s comments on waived testing criteria, see the "On the Issues" section of the CAP Web site’s Advocacy area at www.cap.org.