January 2011
Feature Story

Anne Paxton

It was the study that proponents of home testing of prothrombin time/INR in patients receiving anticoagulation therapy had long awaited. A prospective randomized trial of enough patients to show—perhaps decisively—that weekly home INR testing is superior to monthly clinic INR testing, the currently recommended practice, in reducing the risk of a major event like stroke, major bleeding episode, or death.

But the recently published results of the Home INR Study (known as THINRS) were disconcerting to many. The major finding: Home testing doesn’t appear to bring much improvement over high-quality anticoagulation management.

Published in the Oct. 21, 2010 New England Journal of Medicine, the THINRS study found slight improvements in patient satisfaction and quality of life, and in the average percentage of time the 2,922 patients spent in therapeutic range. The rate of stroke, major bleeding, or death, however, was about the same: 19 percent for those self-testing weekly and 20 percent for those tested monthly at the clinic (Matchar DB, et al. N Engl J Med. 2010;363:1608–1620).

Essentially, there was “no benefit in major clinical outcomes,” says lead study author David Matchar, MD, professor and director, Program of Health Services and Systems Research, Duke–National University of Singapore Graduate Medical School, Singapore. Unable to demonstrate their hypothesis, the authors made the guarded recommendation that self-testing should “be considered for patients whose access to high-quality anticoagulation care is limited by disability, geographic distance, or other factors.”

End of story? Not at all, say Dr. Matchar and other experts in antithrombotic therapy, as well as home testing device manufacturers. They suggest that, for a number of reasons, the THINRS study results may not reflect realities of patient care, and that the trend toward increased home monitoring of INR will continue to improve the treatment of patients at risk of clotting and bleeding.

Some diagnostics industry leaders confess they were taken aback by the findings. From an industry standpoint, “I was disappointed—I thought we would get better results,” says David Phillips, vice president of marketing development for Alere Home Monitoring (formerly Inverness). “I am surprised by the outcome,” says Paul S. Savuto, director of market development and alternate site for home testing device manufacturer ITC. “Intuitively, one would think that testing once a week versus once a month would provide some benefit.”

Although many within the manufacturing side of the business were disappointed with the study outcome, says Phillips, he thinks there’s more to it. He sees the study result as favorable and one the industry can use. “It would have been nice if it had been positive and the outcome said patient self-testing is superior to high-quality anticoagulation management. Salespeople prefer the ‘30-second elevator speech,’” he says—the simple sales pitch that may be their only chance to quickly convince a physician of the value of their product or service. “If you can catch the physician and give your 30-second elevator speech and get them interested, that’s a good thing. THINRS did not give us the quick sound bite, but it did show home testing has advantages—but it will take a little longer to explain why,” he says.

The number of patients on anticoagulation therapy with warfarin in the U.S. is large—about 4.2 million—and steadily growing. According to market researchers, warfarin prescriptions have been climbing about four to 4.5 percent per year since 2006. Of the patients who now receive warfarin and require monitoring, 50 to 60 percent are being followed by private physicians, 30 to 40 percent are managed by an anticoagulation clinic, and so far only about two to three percent are doing home monitoring, says Jack Ansell, MD, chairman of the Department of Medicine at Lenox Hill Hospital in New York City, professor of medicine at New York University, and an adviser to the study.

“The overwhelming majority” of patients on warfarin would be candidates for home monitoring, he believes. But over the 20 years since the first point-of-care coagulation instruments were developed, a complex web of infrastructure, incentives, and insurance factors has stalled what some predicted would be a dramatic shift from clinic testing to home testing.

Dr. Ansell, who helped develop the first point-of-care coagulation instruments, says Medicare policy has limited their wide deployment for home monitoring of warfarin. “The devices were on the market in 1988–1989, but they weren’t certified for reimbursement until 2003. So right there you’ve got 15 years where nobody’s using them.” Europe didn’t have the same obstacle; there the devices are far more popular and widely used.

The majority of patients in this country are managed by their personal physicians with what might be called “routine medical care” or “usual care,” but the currently recommended practice is high-quality clinic testing, Dr. Ansell explains. He founded the Anticoagulation Forum in 1991, a national network of anticoagulation clinics that has grown to include more than 1,500 practices. “Anticoagulation clinics [ACCs] are special programs that do not exist everywhere in the country, and they manage less than 50 percent of warfarin patients, although the actual number is unknown.” (Phillips estimates that the number could be as low as 20 percent.)

Usually ACCs are hospital-based, but sometimes large practices have them as well, Dr. Ansell says. “ACC care has been shown in many studies to provide better outcomes than ‘routine care’ [in general]. That does not mean that every physician is at fault or cannot provide good care. Rather, it is the ‘system’ of management that makes the difference.”

But what the study authors called “the poor record of usual care”—and the fact that high-quality clinic management of patients on warfarin is now the recommended practice—imposed limitations on the THINRS study that significantly affected its findings.

Previous studies comparing home and clinic testing, and a meta-analysis of several studies, had been optimistic about the benefits of patient self-testing, Dr. Matchar says. Why was the THINRS study different? For one thing, it covered a little over a three-year period. “Some of the earlier studies might have been short-term, and some of them had people drop out of the self-testing. We didn’t have that many dropouts. And if you’re following people acutely, you’re generally going to have more adverse events early on, so if there’s a particularly good effect of monitoring, it might be picked up in a short-term study. In our study, which looked at the long-term value of home testing, that effect would tend to get diluted.”

Second, he points out, “patients in previous studies were selected who were especially good candidates for self-testing. In our study, we took anybody on board who was able to use this device, because we were aiming for a very pragmatic study, asking if we were to take a general group, not a selected group, how would they do. It’s possible a very select group would do better.”

Third, and perhaps most significant, was the comparator group. “Our executive committee said we cannot use nondescript care as our comparator,” Dr. Matchar says. “There was an ethical concern since it has been established that some ‘usual care’ may not be of sufficient quality; we couldn’t have the possibility that patients would get low-quality anticoagulation. We needed to ensure that some standard practices were being followed.” These included a responsible party or individual, maintenance of a flow chart, a standard algorithm for management—”basically standard operating procedures,” he says. “In this way, we were not comparing meticulous care with potentially lousy care; the focus was on the difference between home weekly monitoring versus clinic monthly monitoring.

As a result, the comparator group represented what is in the general population a minority of patients, those receiving high-quality clinic monitoring—what Phillips describes as “the best shops in the country.” Says Dr. Matchar, “We weren’t comparing the home-testing group to a random community group—and there is plenty of evidence a community group can be substandard.”

As to cost-effectiveness, home testing proved to be more expensive than clinic testing. “There’s the cost of the devices, the strips, and the training. Over two years, there were fewer scheduled office visits, which are more expensive than phone calls, and there were savings in patient travel. The net extra cost was about $600 a year to do the self-testing,” Dr. Matchar says. In the aggregate, there was an estimated 0.115 quality adjusted life year gained over two years. “So you can say you got an extra tenth of a year of quality of life for $1,200, and that actually turns out to be cost-effective from a societal perspective” (although, he adds, the question of who will pay for that benefit has to be addressed).

A welcome ancillary finding, Dr. Matchar says, was that the majority of people showed adequate competency both for using the device and using the interactive voice-response reporting system, which alerted patients reporting a measurement outside their assigned INR range to contact the study staff. After two sessions, 99 percent of people were able to use the devices, whether they were older like the atrial fibrillation group or younger like the heart valve group. “As far as predicting whether people could use them, there were obvious things like cognitive impairment or stroke, but we couldn’t find anything else. Some very old people were perfectly able to use the device, so we concluded the only way to tell if people can use it is to just have them try.”

Dr. Ansell points to another reason the THINRS study did not show better outcomes: The interpretation of the testing for both the home-testing and clinic-testing groups was performed by the clinic. “So the same person was managing the dosing. That’s immediately going to change the likelihood of the outcomes being different.”

“The problem with anticoagulation therapy is that dose management is generally not performed well in this country, and that’s the key to good therapy,” Dr. Ansell says. “This study did not really address that issue; it only addressed the frequency of doing an INR.” All studies that have compared patients doing home testing and managing their own doses, versus the average physician managing patients, have showed that patient self-management is better, he says. “Overseas, that’s the standard. They’re not only monitoring their own dose; they’re actually adjusting their Coumadin dose, just as a diabetic would for insulin therapy. That really is the ultimate endpoint here and where we should be in this country. But we’re far behind the rest of world in this regard.”

While the Centers for Medicare and Medicaid Services has nothing to do with dose monitoring per se, Dr. Ansell says, cumbersome CMS policy for reimbursing testing and covering the cost of the test instruments is itself a limitation to widespread use of the instruments. ‘It’s a real barrier to getting these instruments out.”

It took Medicare until 2003 to approve payment for home testing devices, and then it was only for the roughly 15 percent of patients receiving warfarin for mechanical heart valves. In 2008, when Medicare approved payment for home testing of patients with atrial fibrillation and venous thromboembolism as diagnoses, that drove the percentage eligible for payment up to 75 percent or 85 percent. However, Phillips notes that it has been only in the past 15 months or so that the new Medicare policy has started to increase home testing.

One major obstacle to wider implementation of home INR testing is posed by a separate Medicare policy. “The POC instruments have been designated by Medicare as a diagnostic test, not a piece of durable medical equipment,” Dr. Ansell points out. “And that’s a very important distinction. A piece of durable medical equipment like a crutch or cane can be purchased from a pharmacy, usually with a prescription, and the insurance company will reimburse for it. These POC monitors are not reimbursable in that regard. They’re reimbursed per test.”

A physician with a patient who wants to do home testing has to lay out the money for the instrument, give it to the patient, “and then every time the patient tests, up to four times per month, you submit a charge to Medicare,” Dr. Ansell says. The CMS also reimburses physicians for training the patient to use the device, but physicians do not recoup their investment in the devices for some time. “You get reimbursed for your outlay test by test. But if you’re going to spend $12,000 or $20,000 or whatever to buy instruments to let, say, a dozen patients do home testing, it will take several years to make back your money.”

Independent diagnostic testing facilities (IDTFs) have emerged, in part, because they have the capital to make such long-term investments practical, Dr. Ansell says. “They purchase the instruments, provide them to patients, train them, collect the results and provide results to physicians, and these companies are able to make back their investments over time.”

“As a manufacturer,” Phillips says, “we provide the monitor and test strips to an IDTF at a volume price.” While there’s nothing precluding practices from becoming an IDTF themselves, there are reasons they tend not to do it, he says. “It’s generally outside their normal routine. What happens if a patient no longer needs the monitor? The practice owns the monitor, so they would most likely want it back. It would need to be refurbished by the practice. How would their patients obtain additional test strips? By mail? Or do they come to the office? Both of these are outside their daily routine. Physicians start to realize the additional logistics of what is necessary, and it becomes less attractive.”

Phillips estimates that only 20 percent to 25 percent of the 4.2 million patients on chronic warfarin are managed by high-quality anticoagulation management clinics. “That means you have 75 to 80 percent of patients being managed by what is termed ‘usual care.’ That tends to be a one- or two-member general practice with only a few warfarin patients. The patients in these practices are the patients who could benefit most from becoming a self-tester.”

The key finding, which the study demonstrates, he says, is that while time in range is modestly better for self-testing, “patients doing self-testing can do it as well as the best run clinics in the country—and the patients like it better.”

Phillips points to an additional feature of the THINRS study data that study co-author Alan Jacobson, MD, presented at an American Heart Association meeting last November. “The average time in target range was 62.4 percent for the high-quality clinics and 66.2 percent for patient self-testing. But when you compare the high and low ranges of the two cohorts, the high-quality anticoagulation clinics ranged from a low of 45.3 percent up to 69.4 percent, while for the patient self-testing arm the time in range was higher: from 52.0 percent up to 80.4 percent of the time. So with many patients, the time in range tends to be significantly higher.”

Phillips says if the time in range in the patient self-testing arm is compared to usual care time in range at 32 percent, as is reported in the literature, the potential improvement in the usual care patient would be significant (Beyth RJ, et al. Ann Intern Med. 2000;133: 687–695) .

Like Dr. Ansell, Phillips is a proponent of patient self-management in addition to self-testing. “It really is like the diabetes model, and if diabetics can do it, there’s no reason why warfarin patients could not perform similarly.” That model shows that patients benefit from taking an active role in their own care, he says. “There are so many things that affect a patient taking warfarin—the patient’s lifestyle, diet, co-morbidities, and concomitant drugs.” During a visit to a clinic in Denmark, he and Dr. Jacobson met several patients who reported that as they tested more frequently, they had begun to notice how different foods or drinks affected their INR levels more than others. “One gentleman found a brand of green tea his wife couldn’t drink because it elevated her INRs, while it had no effect on him. They wouldn’t have known this had they not been able to do their own testing.” Dr. Jacobson’s comment, he says, was that these patients “not only titrate the drug, they titrate their lifestyle.”

In the U.S., by contrast, the educational piece of the anticoagulation therapy puzzle is often missing. “Many patients are simply not aware this technology is available, and that it could be a very helpful option to learn about and actually be doing themselves,” says Lynn B. Oertel, MS, ANP, clinical nurse specialist with the anticoagulation management service at Massachusetts General Hospital and a board member of the Anticoagulation Forum.

She feels that the THINRS trial did add more evidence to further patients’ education. “One thing that came out of the study is that older people, who are a typical demographic for warfarin therapy, are doing it very successfully.” In fact, the authors mention this as information that should reassure clinics and physicians. Oertel’s clinic now has about 220 patients performing self-testing at home. Even very elderly patients with fine-motor issues or visual problems can be candidates, she says, since a family member or caregiver can be trained to perform the fingerstick.

Patients who live in remote rural areas and who might be traveling long distances to a laboratory could find self-testing to be a big relief, Oertel says. In addition, weekly venipunctures on veins can be painful when all that is needed for self-testing with a POC instrument is a fingerstick. And there’s the benefit, too, of consistency: “The patients always use the same device every day, and we know there can be lots of variability among labs.”

Her clinic depends on an IDTF to carry out the patient self-testing program. “Some clinics do the training on their own, as there’s a bona fide billing code designated for that, and they might view it as a way to generate revenue. But our clinic does not have the bandwidth in terms of personnel resources to do that; our staff’s time is fully dedicated to educating and managing our current patient population. So we use the IDTF model to do the training for us. They use a standardized competency checklist to document that the patients are able to perform the test, they are knowledgeable about the care of the equipment, and so on. Once the patient is trained, all the INR results are communicated back to us, and we do all the day-to-day management of the INR values and contacting of the patient.”

From the standpoint of the pharmaco-economics of health care, Oertel strongly believes home testing makes sense. “If people who self-test stay in range more often,” which many studies including THINRS have shown, “then they’re far less likely to develop clotting or bleeding problems that will warrant expensive diagnostic workups or hospitalizations—because someone’s INR went out of range and you didn’t catch it before bleeding occurred.”

What’s the likely impact of the THINRS study? Dr. Matchar has a different take on the study than some observers. “We had a primary outcome measure, which was major events, and we had a sample size that would have allowed us to find a difference that would be clinically meaningful. We didn’t find that difference, so some people would say, ‘Okay, well, based on this study we learned self-testing is not better than good-quality clinic testing; it was a negative study.’ Which is true, from a very restricted perspective.”

But the study turned up no evidence that self-testing harms patients, he points out. “That has been a big issue, certainly for American doctors who, compared with European doctors, seem very uncomfortable with giving the devices to patients and sending them home.” THINRS showed that “if you have some sort of a safety net, like the interactive voice-response system, then you get results at least as good as what you get with a good-quality clinic.” If the financial disincentives to using home devices were eliminated, “then I think American doctors could become comfortable with them.” And based on the range of information the study revealed about ability to use the device, patient satisfaction with home testing, time in target range, and quality of life, home testing also proved worthwhile.

The issues are more complex than may at first be apparent, Dr. Matchar says. “You have to ask, what percentage of patients would not be anticoagulated if we didn’t provide these devices? How many people would be trying to use factor Xa inhibitors, which cost $6 per day? The questions going forward are going to be broader than those just covered by this study.”

That the New England Journal published the home-testing study is impressive, in Phillips’ view. “This isn’t sexy stuff compared to new, novel drugs or innovative implantable devices, but the editors considered it important enough in improving anticoagulation management to publish it. While the results did not support the primary outcomes, however, if you dig into the data, it will help convince physicians that there is a viable alternative to conventional management for the majority of warfarin patients and that properly selected and suitably trained patients can do as good a job as the best managed AC clinics in the country.”

Although important, the THINRS study won’t be the last word on home testing of warfarin patients, Dr. Ansell says. He sees a clear path ahead for home monitoring of INR. “There is no reason why patients can’t manage their own warfarin dose at home, just as patients with diabetes manage their own glucose. It would definitely be better if more people did home testing—and the literature supports that.”