Acontinuum of tests are not ready for clinical use. They vary from tests that will never be adopted in practice to those that show promise but currently lack sufficient evidence of clinical utility. Bill Malone, managing editor of Clinical Laboratory News, interviewed Mike Astion, MD, PhD, medical director of the laboratories at Seattle Children’s Hospital, about how to approach these kinds of tests. Astion is a co-founder of the Patient Centered Laboratory Utilization Guidance Service (PLUGS), a member-based network of more than 100 hospitals and commercial labs whose mission is to improve test utilization.

How did you become interested in this topic?

It is a popular topic among PLUGS members and continuously on the agenda of Seattle Children’s laboratory stewardship committee, which encompasses all subspecialties of laboratory medicine. Just in the last month, this has been a problem in a few domains including autoantibody testing, infectious diseases panels, and genetic testing, for both germline and cancer mutations.

How do you handle a test that is unlikely to ever be ready for prime time?

First, we remove the test from the laboratory formulary so that it is not orderable. In cases where it is the only test offered by a specialty lab or all the testing from that lab is not clinically useful, we remove the lab from the list of approved reference laboratories. Under CLIA, a laboratory medical director is responsible for choosing the reference laboratory used for sendouts, and banning a lab for clinical testing is an example of exercising those responsibilities. I provide an example of a letter that labs can use to notify clinicians that a lab is no longer available (See article online, www.aacc.org/publications/cln).

Is there any way clinicians can still order a test deemed not clinically useful?

In practice, it is difficult to obtain the test. Ordering a test that is not on our formulary is handled like any off-policy decision made inside a hospital. The physician must appeal the decision made by the laboratory medical director, meaning it is escalated to the chief medical officer or equivalent. To have the decision reversed, the physician must provide evidence of the test’s clinical utility and show the test is performed by a CLIA-licensed lab. All that would have to be documented. In general, once a lab is banned, sending out testing to that lab is unlikely, especially if lab leadership has documented their decision-making process.

How do clinicians respond to a test or reference lab being forbidden?

It depends on the case. Clinicians who love a test or lab are unhappy having their testing made unorderable. Others are happy about it. With the ‘Googlification’ of healthcare, patients sometimes get misinformed about the clinical utility of a laboratory test (1). They sometimes develop a fixed, false belief that a test is useful when an unbiased evidence review reveals that it is not.

Increasingly, this leads patients to pressure clinicians for tests the clinician does not want. If we have removed the test from our lab formulary, the clinician can simply tell the patient: “I can’t help you because our lab director has banned the test.” The provider can blame the lab—and that is appropriate. The main responsibility here for laboratory medical directors is to prevent patients from getting falsely diagnosed by useless testing.

How do you respond to people who say that a lab director’s conclusions regarding a set of evidence is an example of “not thinking outside of the box”?

Ah, the dreaded box. Before I tell you what I say—and I have had discussions with providers and patients or their families frequently over many years—let me tell you what I think. If a lab director, because of pressure, sends tests to a lab that produces all kinds of crazy results of questionable quality and a patient gets charged because insurance does not cover these tests, that would be an example of a laboratory medical director who is not thinking at all rather than one who is not thinking outside the box. Laboratory medicine is a boarded specialty and we have reasonable standards. Nevertheless—and I find this unfortunate for all laboratorians—our formularies are replete with obsolete tests, not-ready-for-prime-time tests, and even quackery. This misguided testing is promoted mostly by true believers who are motivated by some combination of wishful or misguided thinking, a desire for fame, or financial pressure. 

What do you actually say in the moment when told your thinking is inside the box?

I say that it is my belief that people come to work to do a great job and that I have never met a provider or patient who purposefully wants to order a test that is not clinically useful. I say that I do not think that the provider is smarter (or dumber) than me or more (or less) ethical. We practice laboratory medicine as a team with a tremendous amount of expertise at our disposal. We often are looking at the case with a different set of experiences and analytic tools. These experiences are abundant. I am not telling the provider or patient that they cannot have the test. I am simply telling them that the patient cannot get the test here at Seattle Children’s. Patients or providers who strongly desire testing will find a way to get it. But we can’t be a party to something that we have concluded produces more harm than good. I finish by saying that I am not asking them to agree that I am right, I am just asking them to conclude that I am trying to be reasonable.

How do you handle an order for a sendout test on the border of being ready for prime time but about which providers and lab leaders have a legitimate difference of opinion?

Our approach to a sendout test is no different than for a new assay we are developing in-house. One successful approach is to perform collaborative research on its clinical utility either as a hospital quality improvement project or as institutional review board-sponsored research. In this collaborative setting, we agree to send out the test and not charge the patient. The protocol—including clinical inclusion criteria, review of results, and whether results could be returned to the medical record—are agreed to by a team of providers and lab medical directors. An example of a memo of understanding to providers that we, and some PLUGS members, use to support this approach is available online at
www.aacc.org/publications/cln.

Who pays for an on the border test?

It depends. Anybody but the patient. It is not fair to charge patients for testing of uncertain clinical utility. In an academic institution or tertiary hospital, we might use research funding from donors or grants. However, if we think the test is particularly promising we cautiously use funding from hospital operations as we would to improve or develop in-house assays.

Why are tests without clear clinical utility so challenging to deal with in practice?

The challenge is that tests, when they are in the transition phase, have less experience associated with them, and the results can mislead clinicians and patients. In the transition phase, less is known about the overlap between disease and health, and about interferences and other technical problems.

If you think a test is approaching prime time, but are not quite sure, why not just declare it a clinical test?

There are a couple of reasons. First, when a test is in a gray area of clinical utility, but it is declared a clinical test, patients are going to pay for it. A number of these tests now cost more than $1,000 and not yet covered by insurance. That is a big, and often surprising, out-of-pocket expense for a patient. Many times, if we disagree with a clinician, we ask them, “Would you like your patient to pay for this?” or “Would you like to put it on a research account?” That reframing often eliminates cases in which providers are ordering tests without a plan that will affect patient management.

References

  1. Astion M. The Google factor: Are the worried well making healthcare sick. Clinical Laboratory News 2014;40(1).
  2. Astion M. Convincing providers and patients to keep testing within your hospital and laboratory’s utilization management system. Clinical Laboratory News 2017;43(4).