Selecting instruments with an eye to CLIA.

According to the regulations, labs can save on salaries if they buy equipment wisely. Of course, there's a catch.

Taking CLIA '88 literally, a lab director who buys a simple-to-use random-access chemistry system and uses it only to run a limited menu of common tests can hire high school graduates to do the testing. The salary savings may be an incentive to buy-but decisions about purchasing equipment involve far more factors than this, especially under the new regs.

At face value, the regulations will reduce salary expenses and ease the staffing shortage. CLIA could be seen to create a market niche for labs that offer only a limited test menu. Referral labs and advocates of bedside testing could expect boom times.

Now comes the tricky part. Suppose the lab manager feels compelled to perform iron-binding capacity (IBC) and high-density lipoprotein (HDL) cholesterol tests, for which the analyzer in question is well suited. Sorry, says CLIA: You can do that only if you hire better-qualified people. So much for saving on salaries.

A simplistic interpretation of CLIA is that it creates a bull market for push-button equipment and modestly qualified staff members. There is yet another twist, however: The laboratory director is now responsible in a personal way for all testing that takes place in the lab, or indeed at any site under the laboratory's control.

Previously, a laboratory could lose its certification if Federal inspectors identified a serious problem. Now, in addition, the lab director and/or owner can be banned under certain conditions from owning or directing a laboratory for a specified time. This change tends to balance the lowered staffing requirements. Few lab directors will lower their personnel standards once they realize they will be held personally accountable for bad PT results. CLIA's personnel requirements represent a floor, not a ceiling.

At best, the regulations allow lab directors to create the best mix of equipment and personnel skills to meet their needs. The rules offer needed flexibility; for example, phlebotomists can be trained to perform simple tests. At worst, laboratorians will fail to pay sufficient attention to the regulations, thereby jeopardizing their certification if inspectors should come to call.

CLIA employs a complexity model that is instrument-specific rather than analyte-specific, as in the past. The model determines both the training that a laboratorian must receive before performing a given test and the type of quality assurance (QA) program that must be in place when such tests are done. For these reasons, the selection of instruments has become inextricably entwined with CLIA's personnel requirements.

In selecting instruments, lab directors have traditionally focused on turnaround time, analyte menu, ease of operation, and the precision and accuracy of methodology used. Issues of cost and workflow have become more prominent in recent years, as has the movement toward bedside testing.

Bedside testing demands ease of operation and a fairly broad analyte menu. The CLIA regulations have both promoted and complicated such testing. Now anyone intending to choose one system over another must take into account the backgrounds of the operators. Are they trained to perform the test? Can they work alone in a satellite facility? Can tests be added without complicating personnel requirements? What's behind these rules, other than the Government's seemingly insatiable need to muddy the laboratory waters? * Site neutral. A basic premise of CLIA is that quality standards should be site neutral. This means that the same standards must be upheld for every individual test, whether it is performed in a physician's office, reference laboratory, emergency room, or central hospital laboratory. The same concept holds true for personnel standards.

According to the CLIA complexity model, simple tests that involve little chance for error can be performed by employees with less formal training than can complex tests with multiple steps. At first the principle appears sound. Flow cytometry testing is an inherently complex process, for example, while dipstick chemical urinalysis is not.

Nevertheless, determining complexity is more difficult than it may appear. For each test, the method of analysis must be considered. This includes not only the steps involved in performing the test but also the complexity of reagents and their preparation. To complicate matters further, CLIA has established a category of waived tests that are not subject to QA or personnel and proficiency standards.(1) * Strata. CLIA declares three levels of complexity: waived tests, moderate-complexity tests, and high-complexity tests. Figure 1 lists six typical generic laboratory analyzers and the level of complexity required to perform specific tests on them.

Exempt tests include those determined by Louis W. Sullivan, M.D., Secretary of Health and Human Services (which oversees CLIA), to pose no reasonable risk of harming patients if performed improperly. Depending on the instrument, a test may fall into two or even all three categories. Consider the CBC. A simple spun hematocrit is a waived (exempt) test. Adding an automated differential makes it moderately complex. If an abnormal smear is reviewed, the test becomes highly complex.

It is expected that 10,000 tests will have been classified very soon - by Sept. 1, 1992. About 75% of this group will be in the moderately complex category. Federal regulators expect exempted tests to satisfy the need of most smaller laboratories, including those in physicians' offices.(1) Any test not on the waived or moderately complex lists is considered highly complex. Why 10,000? The reason is that the list is intended to include every analyte and every instrument commonly used to perform it. Precisely how the Government is conducting these evaluations - a question of deep interest to clinical laboratorians, instrument manufacturers, and many others - is uncertain. The criteria by which tests and instruments are being judged, however, have been revealed. * Complexity. in general, "autonomic" instruments (those that do not require user intervention) have been found to be moderately complex; instruments that require operator activity are highly complex. For example, a barbiturate assay is held to be moderately complex when performed on certain automated systems but complex when done by gas chromatography.

The complexity model assesses each instrument and test under seven criteria (Figure II). For each criterion a score of 1 to 3 (from least to most complex) is assigned. The total score for each analyte on each system is then calculated. Analytes and instruments that score 12 or less are considered to be of moderate complexity. Those scoring 13 to 21 are highly complex. * QA and personnel. Once the complexity level of each test has been determined, CLIA's quality assurance and personnel standards come into play. That may well be the area of greatest sensitivity. Laboratory organizations and individuals have widely criticized the complexity model on the grounds that its lax personnel standards put patients at risk.(2)

QA requirements for moderately complex tests performed according to the manufacturer's instructions (with no modifications) will be less stringent for the first two years after CLIA takes effect on September 1. Starting precisely two years later, the same QA rules will apply to both moderately and highly complex tests, although personnel standards will continue to differ.

No education requirements have been established for employees who perform only waived tests. Lab workers who perform any moderately complex test must have earned at least a high school degree, which for the next five years will also constitute the minimum educational qualification for employees who perform most highly complex tests. After Sept. 1, 1997, however, a minimum of an associate's degree in laboratory science will be required of the latter group. Educational requirements for laboratory directors differ according to the level of complexity and types of testing performed under their supervision. (CLIA's anticipated impact on staffing will be explored in greater detail by Richard B. Passey, Ph.D., in next month's MLO.) * Rising to highest level. If a lab director decides to purchase a system that is considered highly complex, a properly qualified individual must be hired to operate the instrument even if all other tests in the lab fall into the moderately complex category. To the price of the complex system must be added the increased expense of hiring personnel acceptable under CLIA. instrument manufacturers are likely to lobby HCFA, which administers CLIA, to include their analyzers in the less-restrictive moderately complex category. Certainly manufacturers will make efforts to develop machines specifically for that category.

CLIA's personnel stipulation will have special impact on Stat labs. If a lab chooses a moderately complex test menu and instrumentation, the entire testing staff can legally consist of high school graduates. If by virtue of instrument or method selection some assays are rated as highly complex, the lab must hire at least some employees with far more training, presumably at higher salaries. Since Stat labs frequently operate with only one laboratorian, their methods and instruments must be selected with great care to accommodate the level of expertise required by CLIA.

In a further complication, CLIA mandates that a lab in which only waived and moderate-complexity tests are performed, and which staffs accordingly, will be unable to perform certain follow-up tests that laboratorians consider necessary. A good example is an automated differential count on a moderately complex conforming system that is evaluated as requiring a manual slide review. White blood counts are defined as moderately complex only if the analyst is not required to identify abnormal cells. If abnormals are to be screened, however, the screen must be done by an employee authorized to perform highly complex procedures.

Couldn't the ordering physician perform the count? Possibly; under CLIA, the director (if a physician) and other physicians authorized by the director are held to be fully qualified to do all lab tests, including highly complex procedures. Yet this blanket permission is at variance with many state laws. If such state laws are changed over time, directors of Stat labs could theoretically ignore their traditional constituency of technologists and technicians, selecting equipment solely based on ease of operation by physicians - that is, assuming the physicians are willing to do all the lab work.

Equipment design, QC programs, data capture, and audit trails take on greater importance when non-laboratorians perform tests. Each laboratory director should develop a checklist to help determine which characteristics will make the lab's equipment succeed best in the CLIA environment. * Surprises. CLIA rules will have an enormous potential impact on the central laboratory regarding equipment selection and workflow. Since re-outfitting the entire laboratory does not suit the typical lab director's budget, one of the greatest difficulties will involve the dispensation of existing equipment.

Here's an example. By choosing appropriate instrumentation and segregating tests, a lab director makes sure nearly all routine tests are moderately complex. Nevertheless, a few complex assays are needed. Although iron and cholesterol assays are deemed moderately complex, iron-binding capacity and HDL cholesterol tests are complex because they require a pretreatment step. Two years ago, buying a certain expensive chemistry analyzer seemed a good way to contain salary costs, and so it was purchased and installed. It turns out, however, that running a full chemistry profile will demand more highly trained personnel after all.

Similarly, therapeutic drug monitoring (TDM) can be either moderately complex or highly complex (often abbreviated as "complex," just to confuse matters further), depending on the analyzer used. An inattentive lab director might conceivably consolidate TDMs on one of two chemistry analyzers in the laboratory only to pick the one requiring more experienced personnel - and not realize this until examining his lab's certificate. * Distributed testing. The CLIA complexity model will certainly add interest to distributed (point-of-care) testing. That market has been in a state of suspended animation for the past two years as potential instrument buyers wondered whether the regulations would limit or even eliminate such testing. Now that the final regs are out, the implications are clear: Bedside testing has received the go-ahead regarding personnel regulation. Successful instruments in this category will fall under the moderately complex category. To achieve this status for their products, manufacturers will employ microelectronics, biosensors, and solid-state measuring devices that limit the need for human intervention.

More individuals with less specific lab training than before will be able to perform bedside testing. On the other hand, equipment and personnel should accommodate rigorous QC, be able to download data to a central computer system, and include mechanisms that guarantee accuracy. instruments that meet CLIA's requirements in this regard should be welcome in hospital specialty units and elsewhere. Instruments that make it difficult to comply with the regulations are likely to be avoided by wary laboratory directors and managers.

The regs' specific exemption for bedside glucose testing reveals Federal sensitivity to issues of near-patient care. CLIA '88 may serve as the change that moves true time-dependent testing closer to the patient. * What to do. When purchasing equipment, laboratorians should take care not to discount the traditional considerations - turnaround time, analyte menu, ease of operation, and precision and accuracy of methodology - in favor of easy compliance with CLIA. For point-of-care testing, however, CLIA regulations will be the major determining factor in equipment purchase.

CLIA presents the lab director with two options. She may say, "I'll do point-of-care testing, buy easy-to-use equipment, hire less-qualified persons to run it - and put the facility under my license." The lab director has a lot to lose if the facility flouts CLIA rules. Or she can turn to hospital administration and say, "You run the lab; put it under another physician's directorship."

That second option, of course, has a price. While CLIA permits the manager of an emergency room to install a point-of-care analyzer, for example, other appropriate requirements must be met. The bedside testing analyzer will be considered a laboratory in itself and require proficiency testing, QA, and other quality pursuits.

Many clinicians have been clamoring for the right to run their own near-patient facilities. CLIA '88 now makes that prospect easier to achieve. A bachelor's degree qualifies an individual to be the director of a laboratory in which only waivered and/or moderately complex tests are performed - a description that fits nearly all bedside testing configurations.

Near-patient testing will finally get its chance to succeed or fail. No longer can pending Federal regulations be used as an obstacle to implementation. For those who want to institute near-patient testing, the regulations bring opportunity - and increased responsibility.

(1.) Albertson, D CLIA '88 rules go final HCFA details sweeping changes (Washington Report.) MLO 24(4): 17-20, 88, April 1992 (2.) Albertson, D. With more questions than answers, industry studies impact of CLIA '88 (Washington Report). MLO 24(5): 17 18, May 1992

Dr. Lifshitz is director of outpatient laboratories, New York University Medical Center, and clinical associate professor of pathology, NYU School of Medicine, New York City. Dr. De Cresce is director of clinical laboratories, Rush-Presbyterian-St. Luke's Medical Center, and assistant professor of pathology, Rush Medical College, Chicago. The authors publish a monthly newsletter on instrumentation.