College of American Pathologists
Printable Version

  Feature Story


cap today

From Mass General, a value-added coup

October 2001
Karen Titus

Example of patient-specific interpretation for coagulation

Can you stand to hear the words “value added” one more time?

Can you afford not to?

For all that the phrase has been bandied about by laboratory professionals in recent years, rarely is it backed up with what skeptics like to call the cold, hard facts.

But this time it’s different. This time Michael Laposata, MD, PhD, is doing the talking.

Dr. Laposata is a familiar figure to many. He’s director of Clinical Laboratories at Massachusetts General Hospital, Boston, and a professor at Harvard Medical School. He’s a longtime member of the CAP Coagulation Resource Committee and a frequent contributor to CAP TODAY on coagulation-related topics.

He’s also a popular speaker, one whose delivery is as engaging as his message. Those who attended either of his two Executive War College talks this May listened intently to what he had to say. That’s because Dr. Laposata is pulling off what many merely pay lip service to: He’s adding solid value to laboratory testing at junctures where it’s desperately needed—the point where clinicians choose which tests to order and the point at which they assess their clinical significance.

This most fundamental of duties is astonishingly difficult to do, Dr. Laposata notes. "There is a big break in the system right here," he says. "These are the broken links in the chain. But to be able to deliver health care, somehow we have to make sure these elements work."

When, as an MD-PhD student, he sought career advice from a clinical colleague, telling him he was interested in research, the colleague told him, "’Do laboratory medicine. Your entire job is turning on the lights in the morning and being available for an occasional question or two. Hire a good lab manager, and they’ll take care of everything else,’" Dr. Laposata says. "So I entered a residency program in straight lab medicine."

About three months into his residency program, he got a call from a medical resident, someone "who I thought was one of the brighter residents," he says. "We were looking at the patient report, and he says, ’What’s the DAT?’ I thought, if he doesn’t know what a direct antiglobulin test is, how is he transfusing his patients?" When the resident then asked him about a prolonged PTT, it was clear his colleague had "no idea how the factors were arranged."

"I realized then that many of the basic lab test results we [the laboratory] were giving them, they didn’t really understand," he says. When he asked his resident colleague what he was comfortable with, the reply was shockingly brief: electrolytes, the chem 20 panel, and CBCs.

The problem has only gotten worse since that 1983 encounter, as already huge test menus continue to grow exponentially. As Dr. Laposata puts it, "You turn around, and there’s another marker for thrombosis. Primary care physicians cannot maintain current knowledge—it’s just coming at them too fast." Although laboratory information systems and patient and specimen identification processes—not to mention lab instruments and reagents—have improved by leaps and bounds, "We left this out: Did you get the right test in the first place?" says Dr. Laposata.

Too often the answer is a disquieting "no." During the six months each year he serves as an attending at MGH, providing patient-specific narrative interpretations in coagulation, Dr. Laposata says he sees some 50 coagulation cases a day; given the large volume of tests generated at Mass General and by the 70 community hospitals his lab serves, he has a pretty good sense of clinicians’ test choices. "As far as I can tell, many of them—with the absolute best intentions—appear to have a coagulation dartboard in their offices," he says. "The tests ordered often make little sense, and followup discussion with the clinicians reveals significant confusion."

The evidence is not simply anecdotal. A study published in the Dec. 23, 1999 issue of the New England Journal of Medicine noted that 24 percent of primary care physicians and 38 percent of specialists reported that primary care physicians were "in over their heads," Dr. Laposata says, overwhelmed by the changes in the scope of care they were expected to provide.

Clinicians, in short, are in trouble. "They say, ’Look, we need help. This is a big test menu. I heard about this [test] for one second in medical school 25 years ago, and I have no idea how to use it. We’re supposed to pick the six tests to find out if this guy’s genetically predisposed to thrombosis? That’s crazy—please fix this problem by providing more than a big list of tests on a requisition and a pile of numbers generated in the lab.’"

That’s exactly what Dr. Laposata has set out to do. In the seven or so years following his 1983 DAT discussion, "I took notes," Dr. Laposata says. "Then I became director of the clinical labs at the General, and then I greatly impacted people’s promotions and paychecks. And they found it advantageous to follow my lead on this transformation in clinical lab services."

The goal: creating diagnostic rounds, generating narrative reports, and helping clinicians pick tests.

The time is ripe for doing so, and not just at MGH. "Doctors are now in what I call this post-Stalinist Russia phase," he says. "You can go to Moscow, you can criticize Stalin, and nobody’s going to shoot you." Likewise, he says, clinicians can say, "’Guess what? All I know is the chem 20 and the CBC and electrolytes, and I don’t know anything else with confidence.’ And everybody else will nod and say, ’That’s true of me too.’"

Strategy No. 1, he says, has been to trim the test-ordering menu but not eliminate tests. "Let’s face it—the last thing we should do is give doctors a big list of tests."

His laboratory uses reflex testing as much as possible to ensure the right tests are being ordered. In practical terms, "That means the physician says, ’I’ve got a prolonged PTT. I’m not a hematologist, I’m never going to know the eight different tests for the lupus anticoagulant. You’re the expert in the laboratory—you pick the most appropriate tests based on the results of the initial ones. If I check the box to work up the prolonged PTT, you figure the whole thing out with one blood specimen, and don’t keep making me bring the patient back.’"

With that goal in mind, the lab’s coagulation requisition form has been transformed from a three-page list of tests to six boxes that can be checked. "Six boxes," he emphasizes, "one of which is, Prolonged PTT work-up." Checking each box initiates a specific test selection algorithm developed by his laboratory.

Dr. Laposata recognizes pathologists may be uneasy with this approach. However, he has learned the regulations governing locally applied and universal reflex testing and followed them closely in implementing his reflex testing algorithms. Avoiding reflex testing is simply not an option, he maintains. Without an algorithm in place, a prolonged PTT might require three or more patient visits to sort out; likewise, those trying to unravel a prolonged PTT in one patient visit, without using an algorithm, could "spend a fortune on unnecessary lab tests."

Narrative reports are just as crucial as test algorithms. Or at least they should be, says Dr. Laposata.

Certainly in anatomic pathology and radiology, they are. "Imagine if you had a breast biopsy, and you made up a slide and sent it back to the surgeon and said, ’Figure out if it’s benign or malignant, and an anatomic pathologist is sitting there just in case you need him or her.’ Ridiculous," Dr. Laposata says. "Or in radiology, do you think the radiologist is just sitting around saying, ’Just tell us if you have any questions about that complicated CAT scan with a million shadows on it’?

"But what do clinical pathologists do?" he asks. "We give complex batteries of tests, in every area that you can imagine.... And we give them a bunch of numbers, and in an attempt to provide narrative information, a bunch of canned comments under certain numbers. Something we don’t have to think about or process: ’The factor V Leiden is positive—some people get thrombosis.’"

Such standard comments are useless to most physicians, he says. "AP reports are filled with patient-specific information. That’s what we have to do in the clinical lab," Dr. Laposata maintains. It’s up to labs to take the lead, he adds. "Do you think the transplant surgeons are just sitting around, looking at your lab reports and saying to one another, ’One of these days I’m gonna call them and find out about protein S’?"

At MGH, he and his laboratory colleagues "started with the most complicated cases and began working backward toward the simpler ones." What they’re finding, he says, is clinicians desperately need help with interpreting complicated cases. "We haven’t hit the point yet where they say, ’Oh, that’s too simple and the interpretation is not helpful,’" he says.

Dr. Laposata is familiar with the interpretations many laboratories provide, but in one way or another, they fail to supply the needed information. Suspect interpretations include:

  • Those that are only sporadically available. "I’ve seen coagulation interpretations that are valuable, but there’s only one person in the lab who can do it. And when that person goes out of town, patients aren’t allowed to bleed," he says.
  • Interpretations made by someone without expert knowledge. "When the cholesterol is high, the non-expert can say, ’Hypercholesterolemia.’ But that’s it."
  • Those with illegible, handwritten comments.
  • Those that don’t take clinical information into account when it is essential to provide an accurate and informative narrative interpretation.
  • Interpretations confined only to simple cases. Pathologists stymied by complicated cases can’t simply report out numbers and wash their hands of the matter, he says. "You’ve got to be able to do the tough ones."
  • Those that are just sentence-long statements of one-word test results. Dr. Laposata reported seeing so-called interpretations consisting solely of statements such as "The hepatitis test is positive," or "The antithrombin level is low." A sentence, he says, does not, in and of itself, qualify as an interpretation.

What does qualify? Dr. Laposata provided an example from MGH, in which a clinician checks a box instructing the laboratory to work up a prolonged PTT.

Such a report might read like this: The lab first checked to see if heparin was the cause, explaining that after adding an enzyme that degrades heparin, the PTT was still prolonged. Next, a mixing study done on the same sample pointed to an inhibitor. The lab performed the most likely inhibitor test first, the antiphospholipid antibody panel; two tests for the lupus anticoagulant were positive. This finding, along with a normal IgG and slightly elevated IgM, explains the prolonged PTT.

Finally, the report would suggest the possibility of antiphospholipid antibody syndrome and recommend a followup panel be done within six to 12 weeks if the patient’s medical history includes thrombosis, thrombocytopenia, or recurrent fetal loss.

Detailed as it is, such a report can be read in about 30 seconds. "And you’ve got an answer," Dr. Laposata says.

The MGH reports have their origins in the coagulation rounds Dr. Laposata and his colleagues instituted half a dozen years ago. Technologists record laboratory results on a standard data sheet in preparation for residents’ rounds. Prior to rounds, the resident on service collects necessary clinical information. The resident, supervisor, laboratory director, and other clinicians then review all cases. The evaluations are entered into patients’ written and electronic records. Patients are billed for interpretation of the coagulopathy screen, and records of all evaluations are retained in the lab.

Given Dr. Laposata’s expertise in coagulation, that area was a logical one for the MGH lab to tackle first. But other areas have since been added, including toxicology, hemoglobin/anemia, blood transfusion, protein electrophoresis, molecular diagnosis/hepatitis, and autoimmune disease, and more will be added in the future, he says.

Though their efforts were initially aimed only at MGH physicians, Dr. Laposata reported that the MGH coagulation laboratory soon attracted the business of 17 hospitals seeking the narrative interpretations. Other clients—including Mayo Medical New England Laboratory and 50 additional hospitals—followed.

In one survey of users, 58.7 percent of physicians reported that the interpretations shortened the time to a diagnosis, and 71.7 percent said they likely reduced the number of lab tests required to make a diagnosis; an equal percentage of doctors said the interpretations helped them avoid a misdiagnosis. The approval rating for the interpretation service in coagulation was a whopping 98 percent.

Just as useful are the cost savings created by the reports. "It may be clinically valuable, but you have to convince an administrator with numbers," Dr. Laposata says. By his rough estimates, the MGH lab’s approach saves $277 per coag case, or at least half a million dollars annually in coagulation cases alone.

As might be expected, learning the ins and outs of the regulations governing reflex testing has been de rigueur. So, quite frankly, is figuring out how to get paid for the services. But equally important—most important, in fact—is pulling together the key ingredients of a valuable narrative report. "If all of this happens right, you’re significantly improving the quality of care and reducing cost per case," he says.

It’s not difficult for pathologists to establish reflex testing algorithms within current guidelines. Dr. Laposata recalls: "I was once told I should have the doctors order all the tests, 60 tests, say, and then I should then eliminate the 55 that weren’t useful to stay within the guidelines."

Needless to say, Dr. Laposata failed to heed that advice. Instead, he consulted with MGH attorneys. The legal minds concluded that two types of reflex testing exist: institutional, or locally applied; and standard, or universally applied. Examples of the former include mixing studies and thyroid screens; included in the latter category are bacterial cultures, differentials, and urinalyses.

"The most important thing to remember," says Dr. Laposata, "is that any institutional reflex testing must be approved by the institution’s medical policy committee." He also recommended the committee approve the universally applied reflex tests. "I’m going to do that annually as well—I did it for the first time last year," he says. "And there’s no harm in having approval for the standard algorithms also in the minutes of the medical policy committee."

Moving cautiously, Dr. Laposata and his colleagues have developed and gained approval of testing algorithms for nearly every coagulation contingency, as well as other segments of the clinical lab. There’s little doubt the algorithms are sound. Dr. Laposata says: "When I went to the medical policy committee, I said, ’OK folks, what would you like to have us do? I can give you this long list of tests, and you could pick from them, or you can have the reflex test algorithms and check a single box, have only the correct tests done, and draw blood from your patient only once.’ Well, it was approved in about five seconds."

Given the success he’s achieved at MGH, it’s no wonder Dr. Laposata is thinking well beyond Boston. Ultimately, he envisions what he calls a Supreme Court of laboratory medicine experts to provide narrative interpretations, linked to clinicians and pathologists through a Web-based portal.

The possibilities seem endless. But with typical humor, Dr. Laposata suggested two potential denouements. "I figure it’s all either going to come to fruition at some point, or I’m going to die," he says. "One of those things will happen first, and that will stop things."

Karen Titus is CAP TODAY contributing editor and co-managing editor