Feature Story

Bedside bull’s-eye? POC cardiac tests gaining acceptance

April 2002
Anne Paxton

Dramatic confirmation of the diagnostic value of POC cardiac markers has emerged within the past year. Yet how and where such testing is done is determined by the complex relationships between hospital laboratories and POC testing programs.

The need to rapidly triage chest pain patients is widely recognized, and device manufacturers—Dade Behring, Biosite Diagnostics, Spectral Diagnostics, and others—have produced multiple-marker test systems to address this need. "There are several instruments varying from qualitative to semi-quantitative to fully quantitative that are available now for near-patient marker testing," says Kristin Newby, MD, associate professor of medicine in the division of cardiology at Duke University School of Medicine, Durham, NC.

The value of POC cardiac markers has been known since at least 1997. And, during the past year, three significant trials that involved thousands of patients were reported in Circulation and the American Journal of Cardiology. Dr. Newby was director of one of the largest studies, the CHECKMATE trial.¹ All three studies have confirmed that using combinations of troponin I, myoglobin, and creatine kinase at the point of care sharply increases the ability to identify patients with myocardial infarction.

This research adds credence to September 2000 guidelines from the American College of Cardiology and American Hospital Association, which recommended POC testing for cardiac markers. To be clinically useful, the guidelines say, the marker turnaround time should be within 30 minutes—a goal difficult to achieve without using bedside testing.

"These data are all just starting to come to fruition," says Scott Mader, director of professional education and development for Biosite Diagnostics, manufacturer of the Triage Cardiac Panel. "In fact, the maturity of the literature and these clinical guidelines studies, combined with the evolution of technology, has really been driving POC adoption of cardiac markers."

Backing this trend, a review of the ACC/AHA guidelines, published a year later in Lancet (November 2001), stressed that "the most important changes relate to the inclusion of troponin into the risk strategy algorithm." As recently as January 2002, another meta-analysis of the studies reaffirmed "beyond any question that troponin-positive patients are the ones who should be treated with new drugs and would have the highest benefit from it," says Joseph H. Keffer, MD, chief medical officer and vice president of Spectral Diagnostics, a Canadian firm that manufactures the Cardiac STATus panel tests.

Dr. Keffer was director of clinical pa-thology at the University of Texas Southwestern Medical School when he began evaluating troponin and its evolution as a simple, rapid, and proven method of indicating risk of myocardial infarction. "I was interested because it did not require an instrument platform," he says. "Therefore, it was a tool that could discriminate troponin conversion from CK-MB independent of the capital investment limitations that face so many hospitals."

The market dynamic is changing, Mader says, and most new customers who are incorporating Bio-site’s Triage product are using it at the point of care. Improved connectivity has provided an added boost. Bio-site’s system has been on the market for 3 1/2 years and "connectivity in the first two years on the market was a real hindrance," Mader says.

Adding to the growing appeal of the system, he contends, is the usefulness of platform technology that can conduct a variety of assays. "In the ER, you have a variety of patients present with acute conditions, so if you have a system in place to address the broad needs of these patients, it becomes much more functionally valuable at POC versus a single test."

But most hospitals, including one of the sites where the major trials were conducted, have not yet installed cardiac markers at the bedside. Biosite’s Triage system, for example, is now used in about 1,000 hospitals, in some cases for toxicology, in others for cardiac markers. But only a minority of the systems are at the point of care.

Similarly, the 1,000 purchasers of Dade Behring’s Stratus CS locate their analyzers in various sites in the hospital. "If you look at small hospitals, with their emergency departments very close to the laboratory, they don’t need to put their analyzers right in the ED," says Hanjoon Ryu, vice president for market development in cardiac at Dade Behring. But in a hospital with at least 300 beds, more of the installed base is in the ED or in a stat laboratory closer to the bedside than in smaller hospitals.

"Generally, an ED bed is one of the most expensive in the hospital and limited in number," says Mary Lou Gantzer, PhD, president of the American Association for Clinical Chemistry and director of clinical and scientific affairs at Dade Behring. "So the focus is on triaging patients out to radiology or the chest pain unit so you free up the ED bed and get the patient to where the most appropriate treatment is going to happen."

Cardiac markers are a pivotal factor in ED physicians’ decisions about treatment and resource management. "If you’re an ED physician and you have a patient with chest pain, your decision is really threefold," says Dr. Newby. "Do I send him home, hold and observe him, or send him to the cardiac care unit or another inpatient bed? Cardiac marker testing helps. If the initial tests are negative, we might send him to the chest pain unit to just observe. If the marker came back positive, one could make a much quicker decision to send him to the CCU so you’d free up an ER bed quicker, or decide to use a glycoprotein IIb/IIIa inhibitor, or perhaps send the patient to the cath laboratory for an invasive evaluation."

In the CHECKMATE study, which used the Stratus CS, "each marker has a specific profile of when you would expect to first see it in the serum and when it reaches maximum sensitivity for detecting myocardial necrosis," says Dr. Newby. "Myoglobin detects it much earlier than CK-MB or troponin. You’d expect to find it at one to four hours, whereas you wouldn’t expect to detect CK-MB for six hours at the earliest, and probably not until eight or nine hours."

"We said maybe there’s an advantage to putting three different markers together and sampling them at the same time, since they have different profiles and since we know patients come in anywhere from 20 minutes to two days after their symptom onset," adds Dr. Newby. "So if we take advantage of these characteristics of markers which detect earlier versus later, and test them all at one time across the spectrum of patients who are presenting at different times, maybe we’ll have an advantage in detecting patients earlier."

"We combined that philosophy with POC testing," she says, "and asked, Do we identify patients earlier? Do we do a better job of identifying more patients at risk? And if we identify them as marker-positive using this multi-marker strategy, does it correlate with their clinical outcome over 30 days?"

Each site ran its standard protocol for marker testing and CK-MB proved to be the predominant marker test. Under the CHECK-MATE strategy, the sites were given a protocol for testing myoglobin, CK-MB, and troponin at baseline, three hours, six to nine hours, and, if the patient was still under observation, sometime between 12 and 24 hours. A two-marker strategy was also prospectively defined to include CK-MB and troponin, with the question being, How much information does myoglobin provide?

Since the study was completed, many hospitals have adopted troponin as their predominant marker, Dr. Newby says. "It’s becoming the gold standard marker of choice, and there is some testing of myoglobin, but I don’t think it’s still as widely used as we might hope based on this study and previous ones."

"There’s still some work to be done in getting the message out about myoglobin and how helpful it is," Dr. Newby adds.

There are other protocol variations. "The hospitals we worked with predominantly rely on marker testing serially over time, by a schedule similar to the CHECKMATE strategy," says Dr. Newby. "But some hospitals also incorporate myocardial perfusion imaging as an acute test. We don’t do it at Duke, but the Cincinnati site used it."

"When you have a coronary artery blockage," she adds, "you’re not getting blood flow to an area of the heart muscle. If you inject the patient with a radiotracer that’s dependent on blood flow to get to the heart muscle, and there’s obstruction, or what we’d call ischemia, there’s low flow to the heart muscle, and it shows up as an area on the imaging where none of the tracer got to. It’s just another way, like electrocardiograms or cardiac markers, to get an assessment. But it’s not something every hospital can do or can justify based on the volume of chest pain they see."

Another major trial, the 90-minute exclusion study,² demonstrated the value of near-patient testing. It found that troponin and myoglobin can exclude cases of MI within 90 minutes. Study director James McCord, MD, cardiology director of the chest pain unit for the Henry Ford Hospital and Medical Centers, Detroit, initially was interested in a multi-marker approach using myoglobin to rule out MI.

In an earlier study involving 6,000 patients, called the Smart trial, the data suggested that myoglobin and CK-MB combined at zero and one hour identified almost 75 percent of the MIs, or a couple hundred patients in the study. "We said that’s interesting," says Dr. McCord. "If you can identify three-quarters of MIs in an hour, maybe by extending it a little further to 90 minutes or three hours, you could identify a higher percentage."

That hypothesis was borne out. By adding myoglobin to the late-rising markers, the researchers were able to identify 63 of 65 MIs within 90 minutes. "There were only two misses, and they were actually two patients who by their clinical presentation would be identified at high risk anyway, independent of the markers," says Dr. McCord. A third major clinical study published in 2001, the Ninety-Minute Accelerated Critical Pathway,³ also found that POC testing of troponin I, myoglobin, and CK-MB, as part of a standard clinical algorithm, allowed chest pain patients to be triaged accurately within 90 minutes of presenting to the emergency department.

The Henry Ford study was conducted in the background. "The patients had their blood tested at baseline, 90 minutes, three hours, and nine hours, and we had the standard blood tests run in the central laboratory," says Dr. McCord. "Those results were available to clinicians, so clinical decision--making was based on central laboratory data. Then at the bedside, we duplicated the CK-MB and troponin and added myoglobin; that information was not available to the treating physician."

Even though the Henry Ford Health System performed the 90-minute exclusion study, the system does not perform POC cardiac marker testing. The several hospitals and clinics in the system are under a moratorium on new POC testing while standardization is developed. "We have a ton of POC—PTT, glucose, coumadin on outpatients. But as far as cardiac markers, that’s not going to be revisited until our moratorium is lifted," Dr. McCord says.

Cost-effectiveness studies have convincingly documented that in chest pain triage, the enormous expense, and therefore opportunity for savings, relates to avoiding unnecessarily admitting to the hospital patients who are at very low risk, says Dr. Keffer. "Once they’re in the hospital, they all wind up with extensive and expensive further studies, such as angiography, echocardiography, nuclear scans, and so forth." So if a low-risk patient can be managed as an outpatient, dollars are saved and unnecessary testing is avoided.

A 1999 study published in the Canadian Journal of Emergency Medicine supported this assertion. The study reviewed two years of chest pain patient management in the ER at Thunder Bay Regional Hospital, Thunder Bay, Ontario. The findings: The hospital achieved net resource savings in excess of $500,000 per year since the cardiac marker tests were introduced. The director of emergency services reported that the hospital saved approximately $10 of inpatient resources for every $1 invested in POC testing in the emergency room.

Would such savings be unique to Canada’s government-sponsored health system? The reverse is true, Dr. Keffer says. Cardiac marker testing has produced more savings in the United States because the Canadian approach to chest pain is typically much less aggressive and involves a less invasive pattern of care. But Dr. Keffer predicts that will change.

"The American approach is fairly accepted internationally to be more effective as a result of the prospective randomized trial called the TACTICS study, which contrasted conservative management with aggressive, invasive management," says Dr. Keffer. The study⁴ found that improved outcomes are achieved only in patients who are troponin-positive. "So the cost-effectiveness story continues to be reaffirmed," he says. "If you know which patients are experiencing confirmed myocardial necrosis, using objective tools like biochemical markers, you can place your resources in the more appropriate group."

Ryu, of Dade Behring, sees three fundamental customer needs in the POC cardiac marker market. One is assay performance. "They really like to have better sensitivity and precision in cardiac testing," he says. The second is rapid turnaround time. "And the third," he says, "is convenience and quickly getting superior results at the time they want." While these may not be controversial, manufacturers have taken different approaches to reporting results.

Dr. Gantzer, for example, advocates a quantitative test like Dade’s over the qualitative instruments because, she says, it gives a more realistic picture of a patient’s status. And, "since we have all three of the markers, you’re more likely to be able to monitor that patient because each marker has a different time to positivity if the patient is suffering an MI." Patients generally don’t come into the ED at the same point in time into their chest pain, and they can’t always pinpoint when the pain began, she adds, so the clinician can’t be sure which marker will be most useful.

Dr. Keffer, on the other hand, contends that Spectral’s qualitative test is more helpful. "In ER chest pain triage," he says, "the decision is whether the patient is high or low risk. That’s a dichotomous categorization. Once they’re higher risk, then they need to be admitted and hospitalized and followed. And by all means, we recommend that when they’re admitted they get quantitative levels and are followed serially over the next several days because the highest peak is the best indication of the level of risk."

"But you don’t know what the level means at the time you see them in the ER," he adds, "because you don’t know whether it’s on the way up or down." For example, patients sometimes present on the tail end of a new infarct with a new episode of chest pain. "Let’s say the test says it’s a 1 or 2. You don’t know if that’s the peak for that event or whether it’s on its way up to 20 or 100," says Dr. Keffer. "So we simply advocate our use of a qualitative in that initial evaluation of the patient in the ER, because you can get rapid performance and rapid risk stratification."

Regardless of how results are reported, turnaround time in the stat laboratory can be a hindrance to rapid triage. "People in the laboratory will talk about a mean TAT of 45 minutes. That simply means that 50 percent are longer and there’s a significant number that take over 60 minutes," says Craig C. Foreback, PhD, senior lecturer in the clinical laboratory science program at the University of Wisconsin Medical School and a co-author of the -90-minute exclusion study. "So the percentage of time you’re meeting the goal of one hour may be only 75 or 80 percent. What do you do about those patients? That’s where POC testing comes in."

But whether ED personnel can use POC cardiac markers optimally is not yet known. "One thing everybody talks about," says Dr. Foreback, "is when people order tests in the ER, they wait for the results to come back, they do something else, then get sidetracked, and think, oh yeah, I have to check on these results." The initial study at Henry Ford focused on the efficiency of the testing. "The next step would be to see if that improved turnaround time—not just the fact that results are available sooner, but whether they are more readily available to the caregivers right at the patient," he says.

At the University of Wisconsin, no cardiac testing is performed outside the central laboratory, and no one appears to be pushing for it. "So I think it varies from institution to institution," says Dr. Foreback. "And it might have something to do with what people ultimately think can be done with POC in terms of reducing this diagnostic window."

"What we’re finding as a point-of-care company," says Biosite’s Mader, "is the same thing that the investigators are finding, and that is that being able to demonstrate something in clinical trial and being able to put it into clinical practice are two different things. Implementation in the hospital requires the concerted effort of multiple disciplines, and I don’t think the community has been as efficient as it would like to be in encouraging emergency medicine, administration, cardiology, and clinical chemistry to work together to make the processing more efficient."

Dr. Keffer maintains that the laboratory community nationwide has resisted POC testing. "When I was director of laboratories at Parkland Memorial Hospital in Dallas, we instituted about 20 different POC tests," he says. "They were all critically related to key medical decision--making, but we never had a single test that we -didn’t have resistance from the laboratory staff fearing that they would lose their franchise."

"You’ll find it veiled, in various fashions, with the idea that one can do better in the central laboratory, and that was true for years with glucose," he adds. "Everyone knew that fingerstick glucose was not as precise or as sensitive as a laboratory test, but it was only the advance of fingerstick glucoses that permitted tight diabetic control five times a day." The same, he suggests, applies to other tests, such as pediatric strep throat or pregnancy tests before an x-ray—the immediacy of the results tends to outweigh the difference in accuracy.

Resisting change, he notes, is common throughout medicine and is not unique to the laboratory. "We see the same thing among physicians-cardiologists, lots of them, still clinging to CK-MB in making a diagnosis," he says.

Guidelines alone are not enough, Dr. Keffer emphasizes. "I’ve analyzed the literature to date as to guidelines adoption and resistance, including things as notable as the National Institutes of Health consensus conference, which was systematically published over the years in the Journal of the American Medical Association and had essentially no effect."

It’s not shocking that people in medicine are slow to change, he says. "You have to work the system to have them migrate to new paradigms." More has been learned about guidelines in recent years and they now tend to be evidence-based, "not someone’s best estimate but based on solid evidence from prospective, randomized trials when possible," says Dr. Keffer.

More important, he adds, "we now know that to advance medical guidelines, we need to actively disseminate and sell them to clinicians"-in the same way, for example, that the pharmaceutical industry does drug detailing. Inspired by industrial concepts of continuous quality improvement, the most academic of groups in Europe and the United States are now issuing guidelines under special programs, such as the AHA’s Get With the Guidelines, or GWTG. For the ACC, the equivalent is the Guidelines Adoption Program, which is being used to promote better cardiac diagnosis and management.

One of the leading items on the research agenda is improved markers. Says Duke’s Dr. Newby: "We already know, and some of the other recently completed research we’ve done suggests, that we can use the results of troponin testing at least to help us make decisions about using other therapies like glycoprotein IIb/IIIa inhibitors or low-molecular-weight heparin or early invasive testing." The next approach will involve asking, "Does multi-marker testing extend beyond high-risk patients to chest pain units and acute coronary symptoms? Ultimately, we’re looking for even earlier markers of ischemia, before myocardial necrosis occurs. Right now even though myoglobin is early, it’s there only if the heart muscle has died."

Dr. McCord and others would like to see more research on cost-effectiveness. "When the technology first came out, there were several criticisms—one that it wasn’t qualitatively as good as the central laboratory. These now are approved FDA devices, so that’s sort of a moot issue," he says. The only other argument is cost, he adds. "And if you look at the per-unit blood test, it would be more expensive."

"But the big question is, Can POC testing improve the process enough by having faster information to make it monetarily attractive? Some institutions may save an hour, depending on what their turnaround time is, others may save two hours," he says. "No one has implemented the technology and done a good study of before and after."

The other question, he adds, is, Can you improve patient outcomes by identifying high-risk patients and instituting appropriate therapies more quickly? "Most cardiologists will tell you they know of many cases where if they had had the information one or two hours earlier, it would have had a big impact clinically on the patient, but those are anecdotal," he says.

Now that studies have demonstrated the value of POC cardiac markers, researchers need to determine their impact on patient outcomes, says Dade Behring’s Dr. Gantzer. "If you look at the CHECKMATE study, you can certainly identify positive patients much earlier with bedside testing and detect more of them with first measures versus later serial measures. The question then becomes, How rapidly can the ED staff adjust to having results available earlier and really begin triage earlier? The study has demonstrated the potential. Now we want to go beyond it and say, If we start treatment earlier, do we see the improvement we expect in patient outcomes? For example, are they in the hospital less time? Do they suffer less cardiac damage? Do they have less serious sequelae after their admission to the ED?"

Biosite has done a lot of research in this area and has concluded that clinical chemistry is the "linchpin of making POC happen successfully," Mader says. "We find that as testing moves outside the physical domain of the laboratory, the laboratory takes on an even greater role and responsibility in this area." With POC cardiac marker testing, "the only thing that’s really changed is that the speed of results has become much more ’real-time’ for clinicians; in fact, the laboratories’ purview will always be critical to this testing."

"This is why," Mader says, "technology needs to evolve in such a way that LIS connectivity is always accomplishable, that quality assurance is always seamless, and that industry learns to work through the laboratory rather than around the laboratory."

Anne Paxton is a writer in Seattle.