December 2007: Volume 33, Number 12
Outdated Lab Tests
Which Tests Should Be Considered Obsolete?
By Deborah Levenson
A physician’s order for a test that’s no longer useful is a familiar scenario to most laboratorians. Even when newer testing strategies and assays have been proven superior to older ones and are included in clinical guidelines, some physicians cling to the older assays they have relied on in the past. Sometimes these less than useful tests remain in an institution’s protocols or, occasionally, in government guidelines. But running these tests is a waste of time and money because they don’t yield useful diagnostic information, say lab directors. In certain cases, an obsolete test may even impede an accurate diagnosis.
But convincing physicians not to order these tests—while difficult and time-consuming—is ultimately a worthy endeavor, agreed a group of laboratorians interviewed for this article, who emphasized their roles as consultants to physicians and partners in patient care. Explaining to physicians why a certain test is inappropriate is a unique responsibility of the lab, they noted. “We have several thousand tests that are commonly available today, and improvements upon those tests are pretty constant,” explained Brian Jackson, MD. “So there’s lots of obsolescence going on. Physicians can’t keep up with all of this on their own. Labs have to step up to educate them. I don’t think anyone else can do it.”
CLN asked nine clinical chemists to identify lab tests that should no longer be used and queried readers about these tests’ utility in an informal Internet survey (See Figure, below). Tests that generated the most discussion among the laboratorians included two cardiac markers, certain older thyroid tests, and qualitative pregnancy tests. While the lab directors wished that physicians would abandon particular outdated assays in general, they hesitated to call some tests obsolete because these assays are still useful in rare cases.
What’s Really Obsolete?
Completely eliminating most tests that don’t in general offer useful information isn’t advisable, most of the lab directors agreed. They hesitated to label many tests of limited utility as obsolete. “For any of these tests, there are rare cases where you might really need them,” explained Wes Schreiber, MD. “But those instances may be so few and far between that you can easily send the tests off to reference labs.”
Whether a test is truly obsolete is rarely a black and white issue. “When you’re talking about obsolete tests, there’s definitely a continuum. It’s difficult to say who should be the judge of what is completely useless, and individual laboratorians and pathologists are often uncomfortable making that call,” Jackson explained. “Sometimes there’s an unusual but valid use for a test. In general, though, labs are pretty conservative and most could stand to pull some tests off their menus.” That’s just the approach used by Nikola Baumann, PhD. “Calling a test ‘obsolete’ is a strong statement. In my lab, there are tests we don’t perform, but might send out to a larger reference lab,” she commented.
Cardiac Markers: What Can Go Away?
When asked about obsolete tests, many of the clinical chemists immediately thought of assays for cardiac markers: creatine kinase-MB (CKMB) and lactate dehydrogenase (LDH)/ LDH isoenzymes. While the laboratorians all told CLN they agree with recent National Academy for Clinical Biochemistry guidelines that identify troponin as the preferred marker for cardiac injury, most saw a continuing role for CKMB in assessing patients. They advised that LDH was no longer useful because both CKMB and troponin are more specific.
Fred Apple, PhD: “Although troponin is clearly the superior marker, some labs in developing countries can’t afford it. CKMB and total CK have roles as cost-effective options in places that can’t afford troponin.”
Catherine Hammett-Stabler, PhD: “There are places where troponin isn’t readily available. You should certainly use troponin when available, but when it’s not, CKMB is acceptable. I don’t think you can get rid of CKMB any time soon in the U.S. because some cardiologists and emergency room (ER) physicians often aren’t comfortable using a single troponin test, especially troponin T. But using troponin and CKMB together can be a problem. For example, patients with renal dysfunction sometimes have positive troponin T without a clear increase in CKMB. Some cardiologists and nephrologists think the situation could mean necrosis, but it may not be indicative of MI.”
Nikola Baumann, PhD: “In our hospital, physicians routinely order both CKMB and troponin I, which is unnecessary. Most of the time, CKMB will agree with the troponin I result for the purpose of diagnosing a myocardial infarction. But there may be times when the troponin I and the CKMB results do not agree. I always wonder how clinicians react to those discrepant results and whether they appreciate that troponin I is the more sensitive and specific test.”
Wes Schreiber, MD: “Troponin has better sensitivity and specificity than CKMB, but CKMB is still a very good test. There’s a place for CKMB in determining if a patient has had a second myocardial infarction (MI) following a recent MI. That’s because CKMB drops more rapidly than troponin. Although CKMB isn’t the preferred test to do when you suspect an MI, I’d be reluctant to take it off a lab test menu.”
Nader Rifai, PhD: “Children’s Hospital Boston uses troponin to identify minor myocardial injury resulting from a procedure or a drug’s cardiotoxic effect. For this purpose, CKMB may be misleading, and troponin is clearly the way to go.”
Bette Seamonds, PhD: “Troponin is really now the gold standard. A few hospital labs have quit offering CKMB. It’s superfluous. I’m trying to convince our ER physicians and cardiologists that it should go. They argue that CKMB has a use in identifying renal patients with muscle damage.”
Lawrence A. Kaplan, PhD: “We dropped CKMB from our testing menu at Bellevue Hospital (New York, N.Y.). It’s rarely justified, so I rarely allowed it to happen. Some physicians argued that CKMB could size a myocardial infarction, but I never saw the data.”
LDH/ LDH Isoenzymes
Lottie Goldsmith, BS, CLD: “The total enzymes, like LDH and AST, were used thirty to forty years ago to indicate coronary damage. Now we have tests that are so much better.”
Catherine Hammett-Stabler, PhD: “Years ago LDH was a great cardiac marker for those MI patients who came in a long time after symptoms began. It was the last cardiac marker that got evaluated, and it stayed around for a long time. But with troponin, it’s simply not needed now.”
Wes Schreiber, MD: “Lactate dehyrogenase is very nonspecific. It’s frequently abnormal because it is found in every organ, so as a cardiac marker, you shouldn’t even think about it. LDH isoenzymes have been made outdated by troponin.”
Nader Rifai, PhD: “I haven’t seen that test for 20 years.”
Is ESR Still Useful?
An inflammatory marker, erythrocyte sedimentation rate (ESR) is a nonspecific marker of inflammation that typically has been used in conjunction with other tests to help diagnose conditions associated with acute and chronic inflammation, including infections, cancers, and autoimmune diseases. The clinical chemists agreed that ESR, while quick and easy, is of limited use.
Bette Seamonds, PhD: “In the 1970s, ESR was part of a battery of inflammatory marker tests. Today, it’s like voodoo when you have high-sensitivity CRP. Some protocols actually still call for it for stroke. But why do ESR in a stroke situation, when you want a rapid result? You can do CRP in minutes, while ESR takes an hour.”
Catherine Hammett-Stabler, PhD: “For most inflammation, CRP and other markers are better. A couple of papers from 1999 do justify ESR to identify temporal arteritis, a condition related to stroke.”
Nikola Baumann, PhD: “Everyone agrees it’s completely nonspecific but it seems that few people want to get rid of it, and physicians continue to order it. An elevated ESR may reassure what the physician already suspects.”
Wes Schreiber, MD: “ESR is very cheap and very easy, but it’s also very nonspecific. Most people considered it obsolete ten years ago.”
TSH Supplants Free Thyroxine (T4) Index and T3 Uptake
The group of lab directors generally agreed that the free T4 index and T3 uptake no longer belong in hospital labs. That’s because the tests estimate hormone levels in the blood and have been replaced by the combination of more sensitive thyroid stimulating hormone (TSH) assays and free thyroxine assays that give direct measurements. However, they may have a valid place in reference labs because the free T4 index can be useful for a certain thyroid condition in newborns.
Brian Jackson, MD: “This is a pretty clear- cut case of something being really obsolete. The free T4 index is an estimate that requires a calculation. It became obsolete as soon as good free T4 assays became available.”
Wes Schreiber, MD: “TSH is now the best test for thyroid function. T3 uptake was developed as an indirect way to estimate the amount of thyroxine-binding globulin, which affects the level of free thyroxine. Now, TSH can determine both hyperthyroidism and hypothyroidism very well.”
Catherine Hammett-Stabler, PhD: “These are older tests developed to estimate free T4 in serum. T3 uptake should be replaced by one of the free T4 assays, but I still have endocrinologists who are upset with me for getting rid of T3 uptake. Manufacturers would do the world a service by stopping production of T3 uptake assays.”
Lawrence A. Kaplan, PhD: “In most cases, I think it’s useless. But there are rare cases of newborns with low levels of thyroid binding globulin that T3 uptake can pick up. I made this test available through a reference laboratory just for these rare cases, though.”
Bette Seamonds, PhD: “I want to get rid of this test in the worst way. T3 uptake is one of God’s most useless tests. But it makes money. Half of my workload is outpatient, so I see a lot of requests for this test, although not as much as in the past.”
Why Use a Qualitative Serum Pregnancy Test?
While all of the experts agreed that quantitative serum pregnancy tests are far superior to qualitative ones, most agreed that eliminating these urine tests from the ER would be difficult because some situations require a quick determination of pregnancy. Some of the laboratorians offered strategies for balancing accuracy and speed.
Nikola Baumann, PhD: “I could buy into the idea that qualitative pregnancy tests are obsolete in the central lab setting. We have good quantitative serum assays that have better sensitivity than the qualitative tests. However, I don’t think that point-of-care urine pregnancy tests will go away because the test is noninvasive, quick, and easy. In an ER setting, the qualitative urine test gives the fastest answer as to whether the patient is pregnant or not. In a STAT lab, serum assays could easily be validated for urine and used for the same purpose.”
Catherine Hammett-Stabler, PhD: “The qualitative test is a headache because it’s not foolproof. But you need it in some situations, such as if a woman needs x-rays.”
Bette Seamonds, PhD: “We have analyzers for this test on urine in the ER. It’s hard to get accurate results manually because the timing means the test will likely be done incorrectly. The qualitative test done on whole blood is different story, if there’s not interference from hemolysis. I have an analyzer in the chemistry lab that gives results in 15 minutes. It’s the best option if the ER gets the sample to me quickly.”
Lawrence A. Kaplan, PhD: “This very rapid test is needed in the ER, although you could go another route. Get the serum in the laboratory and run it as a yes/no qualitative test.”
Lecithin/Sphingomyelin (L/S) Ratio: Still Useful?
Another controversial test is L/S ratio. A marker for fetal lung maturity that helps predict whether a baby will be born with respiratory distress syndrome, L/S ratio wasn’t considered useless by the laboratorians, but neither was it their first choice for this purpose.
Nikola Baumann, PhD: “Instead of L/S ratio, many labs perform the Fetal Lung Maturity (FLM) II (Abbott , Abbott Park, Ill.) test as the first-line test. If you get an intermediate or immature result using FLM II, one can do a secondary test. That could be the L/S ratio or lamellar body count, depending on the lab.”
Catherine Hammett-Stabler, PhD: “This test is very cumbersome. It takes hours to perform, it’s not very specific, and there are hundreds of variations on the original L/S ratio method. Lamellar body counts or FLM are better.”
Bette Seamonds, PhD: “When my hospital had maternity services, physicians who wanted L/S ratio had to speak to me directly. Instead, we used FLM in combination with fetal fibronectin (fFN). fFN gives a sense of whether the mom is high-risk for prematurity. Combined with FLM, it’s a pretty powerful tool.”
Changing Physician Ordering Habits
How should laboratorians convince physicians to use better tests? The group suggested careful explanations, educating interns, and reports that point out poor test choices to physicians.
Nader Rifai, PhD: “Remember who you are and what you are there for. You want to be a consultant, not a clerk. Contact whoever is ordering the test and explain why he or she should discontinue it in terms of good practice and cost restraint.”
Catherine Hammett-Stabler, PhD: “We have several ways these are handled. In many cases, our staff in Referral Testing does the initial investigation as to why the test was ordered and discusses the issue with the physicians. In other cases, the resident or fellow follows up with the ordering physician, discussing the testing choice and why another test is better. The most success in changing ordering habits comes from direct conversations. Sometimes a physician wants the test no matter what. In these cases, we call the results to the physician and talk about what the result doesn’t mean and steer him or her to another test. We may need two or three rounds with ‘repeat offenders.’ ”
Bette Seamonds, PhD: “I beat on my interns. I tell them what not to order. As a clinical chemist, I can raise them to a different level.”
Lawrence A. Kaplan, PhD: “It’s difficult to convince physicians to change, but it’s ultimately ‘doable.’ The most inflexible are the private practice physicians. In the hospitals, the doctors are under capitation. You tell an in-house physician, ‘Stop, you’re costing me too much money.’ When private practice doctors order tests, both they and the lab make money.”
Brian Jackson, MD: “Reference labs are in a difficult position to directly challenge physicians’ orders because we have no direct relationship with those physicians. On the other hand, academically-affiliated reference labs can be an educational resource for other laboratories. As part of ARUP’s Analyzing Test Ordering Patterns (ATOP) program, we take a broad look at our clients’ orders and provide reports that identify obsolete tests and other potentially misused tests. We also provide literature references the pathologists can take back to their physicians. Otherwise, reference labs are in a difficult position for changing what physicians order because we have no contact with them.”
Wes Schreiber, MD: “As a rule, formal education programs have only short-term effects. The best tactic for changing physician behavior is to take the tests off the menu. Some of these tests should be discontinued altogether, while others should be available only by consultation with the lab.”
Other Outdated Lab Tests
While the group of lab directors most frequently identified old cardiac markers as outdated tests, they considered several other assays of very minimal use or obsolete. One is prostatic acid phosphatase, which has been replaced by PSA, a much more sensitive and specific marker of cancer. While two lab directors noted that some hospitals still use prostatic acid phosphatase in sexual assault workups, others pointed out that PSA is just as useful for this purpose.
Zinc protoporphyrin (ZPP), abandoned as a lead poisoning test following CDC’s 1991 redefinition of lead poisoning to a cutoff level below what the assay can detect, is another test that is clearly unwarranted in most cases. However, some said ZPP could be useful in reference labs for spotting iron deficiency or whether high lead levels reflect a chronic rather than a single, acute exposure. Likewise, the lab directors agreed there is a place for parathyroid hormone-related peptide in reference labs for some tumors. The group roundly criticized the Schilling test as an outdated means of testing for vitamin B12 because it involves radioactivity.
In thinking about old tests and what has replaced them, many lab directors noted that assays are not only becoming more sophisticated, but are also enhancing laboratorians’ role as partners in patient care. Fred Apple, PhD explained the factors driving these trends. “Assays are able to measure analytes at lower and lower concentrations, with improved precision, and are now directed at more tissue specific protein markers, with fewer false analytical positives. There’s been a shift toward measuring analytes in their more physiologically active form and towards molecular testing. We’re getting better at measuring low levels of biomarkers that are more organ-specific. We’ve moved away from qualitative tests to quantitative ones, from activity to mass measurements, from total to free fractions, and toward more physiologically active forms that better correlate with the pathophysiology of disease. It’s an exciting time in laboratory medicine.”
Discussing the Tests
These nine clinical chemists shared their views on outdated tests.
Fred S. Apple PhD
Professor of Laboratory Medicine, Department of Laboratory Medicine and Pathology, University of Minnesota School of Medicine, Minneapolis, Minn.
Medical Director of Clinical Laboratories and the Clinical Chemistry and Toxicology Laboratories at Hennepin County Medical Center, Minneapolis, Minn.
Nikola Baumann, PhD
Director of Clinical Chemistry, University of Illinois Medical Center at Chicago
Assistant Professor, Pathology, University of Illinois at Chicago
Member, CLN Board of Editors
Lottie Goldsmith, BS, CLD
- Co-Director, Transplant Laboratory, University of Miami Medical Center, Miami, Fla.
Catherine Hammett-Stabler, PhD
- Associate Director, Department of Pathology and Laboratory Medicine, University of North Carolina (UNC), Chapel Hill
- Associate Director of Core Laboratories, UNC Hospitals and Clinics
Hammett-Stabler has received funding to study FLM and LDH isoenzymes in the past.
Brian Jackson, MD
- Medical Director of Infomatics, ARUP Laboratories, Salt Lake City, Utah
- Assistant Professor of Pathology (Clinical), University of Utah, Salt Lake City
Lawrence A. Kaplan, PhD
- Consultant Clinical Chemist
- Director, Clinical Chemistry, Children’s Hospital Boston (Mass).
- Professor, Harvard Medical School, Boston, Mass.
- Editor, Clinical Chemistry
Bette Seamonds, PhD
- Clinical Chemist and Director of Point-of-Care Testing Services, Mercy Health Laboratories, Darby, Pa.
Wes Schreiber, MD
- Consultant Pathologist, Vancouver General Hospital, Vancouver, Canada
- Professor, Department of Pathology & Laboratory Medicine, The University of British Columbia, Vancouver, Canada
- Member, Editorial Review Board, Lab Tests Online, representing American Society for Clinical Pathology