Struggling with warfarin initiation and dosing
An old standby suffers from a lack of standardization
Keywords: Hospital physicians struggle to standardize warfarin dosing and initiation
by Bonnie Darves
Published in the December 2012 issue of Today's Hospitalist
FOR YEARS, physicians waited for the arrival of a new breed of anticoagulants that would eliminate many of the problems associated with warfarin. While those drugs were approved two years ago, the reality is that physicians—particularly hospitalists—continue to struggle with anticoagulation.
Although the FDA estimates that more than a half million patients are using oral anticoagulants like dabigatran and rivaroxaban, the vast majority of patients continue to use warfarin. And when managing patients on warfarin, hospitalists say their job is complicated by the fact that there's little standardization about when prophylactic anticoagulation
should be started, when it should be stopped and what warfarin dose should be used.
|"One of the top reasons we admit patients is either too much Coumadin or too little."|
–Tamer Mahrous, MD
Inpatient Physician Associates
Many surgeons, specialists and even primary care physicians seem to follow their own rules when it comes to anticoagulating patients with warfarin. When problems occur, it's typically hospitalists—not the physician who wrote the script in the first place—who have to rush in and pick up the pieces.
Hospitalists who are still all over the map with warfarin initiation might think that the solution is to prescribe new oral anticoagulants. But for physicians treating bleeds in those patients, the situation can be even more challenging. (See "New drugs, new problems.") Because there is no proven way to reverse the effects of those oral anticoagulants, hospitalists face even more challenges when these patients hit the wards.
The bottom line is that hospitalists managing anticoagulation often feel like they're between a rock and a hard place. "If the patient has a bleed, there will be an investigation asking us why we gave the patient anticoagulation," says Margaret Fang, MD, a hospitalist and medical director of the anticoagulation clinic at the University of California, San Francisco. "And if they have a clot, we'll be asked why we didn't use anticoagulation."
The Wild West
As all hospitalists know, lack of standardization in how warfarin is prescribed as an anticoagulant is a big problem, particularly at initiation and in perioperative bridging.
According to Aynna Yee, MD, director of the hospitalist program at Santa Rosa Memorial Hospital in Santa Rosa, Calif., and a partner in the privately owned group Golden Gate Hospitalists, the challenge is trying to balance many individual patient factors. "Patients and their situations are all so different," Dr. Yee explains. "Do they have atrial fibrillation, a mechanical valve? The type of surgery or procedure is also a factor when you're weighing bleed risk vs. thrombotic risk."
Her hospitalist group is spearheading an effort to standardize the use of warfarin in the hospital. "We're working on a pharmacy-driven protocol so that it's easier for physicians who aren't as familiar with warfarin or other anticoagulants to order them," Dr. Yee says. "And we're trying to get some agreement on when warfarin should be stopped before surgery because it's key for hospitalists to be involved in that decision."
But even low-risk anticoagulation cases present challenges. Hospitalist Tamer Mahrous, MD, a partner in the privately-owned Inpatient Physician Associates in Lincoln, Neb., would be happy if fewer patients were admitted with outpatient anticoagulation complications. As far he is concerned, it's still the Wild West out there, while the number of patients who are chronically anticoagulated, regardless of whether it's warranted, is rising.
"One of the top reasons we admit patients is for anticoagulation-associated drug complications—either too much Coumadin or too little," says Dr. Mahrous, noting that up to a half dozen of his roughly 14 patients a day are on warfarin.
"Coumadin is erratic, and it interacts with so many drugs." Leading the list of problem interactions that Dr. Mahrous' group sees are quinolones, Bactrim and amiodarone.
Here's a typical scenario: A nursing-home patient has an abnormal urine. Because it's the weekend, the nurse calls the covering physician, who orders Bactrim without knowing (or remembering) that the patient is on warfarin.
"Often, a lot of people are involved in the care, and these things get out of hand," Dr. Mahrous explains. "And then patients are in our ER with a big bruise or a bleed, and the hospitalist is admitting them for reversal."
Problems with chronic anticoagulation
In other cases, patients on chronic anticoagulation are inadequately managed in the outpatient setting, Dr. Mahrous adds, or are admitted for an elective procedure without a solid (much less standardized) plan for initiating anticoagulation. Here's one of his typical patients: an 85-year-old admitted for hip fracture repair. Among surgeons, Dr. Mahrous has seen orders for starting or bolus doses of warfarin anywhere from 2 milligrams to 10 milligrams, as well as differing approaches to using rivaroxaban, Lovenox or subcutaneous heparin.
"I still see a tremendous amount of variation from doctor to doctor and from specialty to specialty," Dr. Mahrous says. "It seems to depend on who's rounding, how they were trained and maybe even how they're feeling that day."
He notes that there's no protocol in place at the hospital where he works for drug dosing. And "we aren't really pushing yet for standardization because we've got so many other care issues to deal with, with a new EHR coming online." While the hospitalists follow protocols for monitoring anticoagulation, as mandated by the Joint Commission, they don't have any for dosing initiation or daily adjustments.
But even if the hospitalists standardize their approach to anticoagulation, Dr. Mahrous says there may still be problems in the way that physicians outside the hospital order warfarin.
"When I came out of training, everyone started patients on 5 milligrams," he explains. "But with a frail 90-year-old, I am not going to start her on 5. My experience after more than a decade as a hospitalist is that if we start everyone on 5 milligrams, we will have a lot of people overshooting right off the bat. And then you end up holding the dose."
Hospitalist Manoj Mathew, MD, regional lead physician for the California mega-group HealthCare Partners, which has 110 hospitalists, says he similarly sees wide variation among how physicians order anticoagulation. That's particularly true in the perioperative setting.
"Even if the hospital in question has an internal protocol for anticoagulation initiation," he explains, "we see significant variation out there, especially with warfarin." And "a lot of dosing variation with warfarin is subjective, not protocol-driven."
One surgeon, for example, might start patients on 10 milligrams, while another one goes with 5. To reduce the variation, the hospitalists with HealthCare Partners have taken an increasingly proactive role in advising on or assuming responsibility for anticoagulation initiation.
"At this point, the surgeons are pretty much letting us do what we do with anticoagulation bridging, because we've built up a certain level of confidence and trust with them," Dr. Mathew says. He also notes that pharmacists, both in the hospital and the warfarin clinic, manage dose adjustments.
Harry Rosen, MD, a hospitalist at West Hills Hospital & Medical Center in West Hills, Calif., has taken a similar approach with surgeons because of the variation in warfarin orders he has seen.
"Some orthopedic surgeons do something that's guideline-approved but not appropriate," he notes. "They might start the patient on low-dose warfarin, but they don't recognize that it takes five to seven days to kick in. So patients are completely unprotected for the first several days."
In such situations, Dr. Rosen will often propose a hybrid strategy that bridges the gap in multiple ways: adding a faster-acting anticoagulant like subcutaneous heparin, Lovenox or rivaroxaban. "I rarely have a problem with the surgeon when I want to add something on top of Coumadin," he says. "And at least I know that the patient is better protected."
Lack of indications
At the other end of the spectrum, hospitalists see situations in which patients have no clear indication for being on anticoagulation. That happened recently to Dr. Yee when a woman in her early 60s was admitted for pneumonia and had been on warfarin for five years for a trauma-related thromboembolism.
"I called the patient's primary care physician, and she couldn't give me a good reason why her patient was still on warfarin," Dr. Yee recalls. "She even admitted she'd done no follow-up testing, but she insisted her patient stay on it because she 'had a gut feeling' the patient was at risk for another event."
Even worse, hospitalists' patients may be both inappropriately anticoagulated and clueless about why they're on the drugs. "We have a lot of patients who come in on warfarin and have no idea why they're on it or even what an INR is," Dr. Mahrous reports. "So we're left with trying to figure out why."
It often turns out that the patient had a one-time DVT or another transient event. "When that happens," Dr. Mahrous says, "I try to get them off of anticoagulation. Because my experience tells me that at some point, almost everyone on these drugs will have a bleeding complication."
Bonnie Darves is a freelance health care writer based in Seattle.
New drugs, new problems
THE CHALLENGES that hospitalists face in anticoagulation have been ratcheted up by the introduction of oral anticoagulants like dabigatran and rivaroxaban. While the drugs may be easier to manage than warfarin because they don't require frequent monitoring, hospitalists who have had to treat bleeding from oral anticoagulants have a decidedly different perspective.
"We see only the complications of these newer anticoagulants, not the patients who are doing well on them," says Sandeep Sachdeva, MD, a hospitalist and researcher at Swedish Medical Center in Seattle. "The problem is that we don't have an established reversal strategy for those bleeds besides supporting the patients and waiting it out."
That waiting game can get hairy, says Thomas Rivers, PharmD, clinical lead pharmacist at Swedish. "When patients on these newer agents do bleed, it can be very impressive, and that's a big area of concern now for the hospitalists who get these patients on their service," he says. "They are asking, 'What do I do now?'"
And because there's no surefire way to measure the anticoagulation effect of these new agents, perioperative planning is challenging as well.
"If patients have impaired kidney function," says hospitalist Margaret Fang, MD, medical director of the anticoagulation clinic at the University of California, San Francisco, "you get residual levels of anticoagulants, and it's very hard to figure out how much there is with these new agents. This is what we often get consulted on, and much of what we recommend is a best guess or an educated recommendation."
Using prothrombin complex concentrates (PCC) or fresh frozen plasma (FPP) in concert with saline and transfusion is emerging as an appropriate reversal strategy for dabigatran. But it's neither guaranteed nor evidence-based, Dr. Rivers notes, and there's no established dosing strategy for those products.
He cites a recent case: a patient whose renal function took a nosedive in the several months after her physician had prescribed dabigatran. The patient had not seen her physician, and when she landed at Swedish, she had an INR above 8. That put her, he says, "at exceptionally high risk of having a life-threatening bleed."
"For the INR to get that high, she had to have a huge amount of drug in her body," Dr. Rivers explains. "If it had been warfarin, I estimate that the INR would have been in the 15 range."
That incident and others like it prompted Swedish to form a work group to develop both monitoring and reversal strategies for the new anticoagulants. "What we're finding out is that a lot of physicians take a 'set it and forget it' approach to prescribing these agents, because they've been told that they can just start them and see the patients in six months or a year," Dr. Rivers says. "But like warfarin and heparin, these are extremely powerful anticoagulants, and it doesn't take a whole lot to get into real trouble with them."
Toby Trujillo, PharmD, director of the inpatient anticoagulation service at the University of Colorado in Denver, agrees that while the new agents offer a good alternative to warfarin for many patients, prescribers don't necessarily have a handle on the differences between the two types of therapies beyond monitoring.
"These agents come with their own challenges," Dr. Trujillo says, "and some physicians don't realize that they still have to worry about interactions." Doctors also need to remember that information on drug-drug interactions evolves over time.
"If a patient comes into the hospital with a problem and you do a little digging," Dr. Trujillo explains, "you may find it's a drug interaction that hasn't been described yet."
New guidelines, but will they help?
HOSPITALISTS LOOKING TO GUIDELINES to help them sort out thorny anticoagulation issues won't find much comfort. The American College of Chest Physicians (ACCP) published revised guidelines for antithrombotic therapy in the February 2012 issue of Chest. But especially in the area of DVT and PE prevention, the revisions either tossed out or substantially downgraded much of the evidence on which previous guidelines were based.
The new guidelines, for instance, call for being more selective when deciding who's a candidate for DVT prophylaxis. The guidelines also put aspirin back into the picture as a viable option (if not the preferred one) for DVT/VTE prevention in major orthopedic surgery. They also call for taking patient preferences more fully into account when recommending an antithrombotic strategy.
"Where previous ACCP guidelines recommended more aggressive anticoagulation, many of those recommendations have been changed to 'weak' for low-risk patients," says Margaret Fang, MD, a hospitalist and medical director of the anticoagulation clinic at the University of California, San Francisco. "And for the moderate-risk patients, they've basically said, 'we don't know.'"
Steven Deitelzweig, MD, system chair of hospital medicine for Ochsner Health System in New Orleans and co-author of the 2009 "Contemporary Hospitalists' Guide to Anticoagulation," quantifies the extent of the changes.
"In the 2008 guidelines, there were 182 grade 1A recommendations for clear clinical benefits of anticoagulation. Now there are only 29," says Dr. Deitelzweig, who is also professor at the University of Queensland School of Medicine. "What the panelists discovered—and many of the chapter leaders were methodologists this time—was that the literature from the past was not as good as we thought."
Of particular importance to hospitalists, Dr. Deitelzweig adds that the guidelines don't take a clear-cut approach to DVT prevention. This comes in an era when the Joint Commission and the Centers for Medicare and Medicaid Services are pressing hospitals to be more vigilant in DVT prevention and anticoagulation oversight. "What the panelists focused on this time, unlike in the previous guidelines, was a system-based approach to thrombosis," Dr. Deitelzweig notes. The revised guidelines also include recommendations for using newer anticoagulants for atrial fibrillation and certain major orthopedic surgeries.
The guidelines do recommend anticoagulation for proximal leg DVT and symptomatic distal DVT, but take a wait-and-see approach for asymptomatic distal DVTs. They also advise a far shorter period of anticoagulation—three months, not six to 12—for DVTs associated with surgeries or mild risk factors.
In the case of PEs, the guidelines advise anticoagulation early on but no routine use of thrombolytics unless the PE is large or symptomatic enough to produce major discomfort. In the first few days after PE onset or diagnosis, the guidelines recommend using once daily dalteparin, tinzaparin or fondaparinux, or twice daily enoxaparin, followed by warfarin. And as with DVTs, the recommended treatment duration for PEs associated with surgery or minor risk factors is three months, not six to 12 months.
"There's really been a pendulum swing away from giving DVT prophylaxis to everyone," Dr. Fang points out. "People are starting to say, 'We should do DVT prophylaxis, but only in the people who are going to have the most benefit.'" But that raises another challenge: The risk-stratification tools in use are cumbersome "and difficult to remember," she says. The tools that Dr. Deitelzweig recommends include the Caprini score, the Padua prediction score and the Pulmonary Embolism Severity Index.
Pharmacists to the rescue—sometimes
BECAUSE OF THE INCREASING COMPLEXITIES of initiating and monitoring (and sometimes reversing) inpatient anticoagulation, many hospitals are putting clinical pharmacists on the front line. At Santa Rosa Memorial Hospital in Santa Rosa, Calif., that move has gone a long way toward standardizing anticoagulation monitoring and management, explains Joanne Bell, PharmD, the hospital’s clinical pharmacy supervisor.
“Our hospitalists know they can write for us to handle the dosing and the monitoring, and that takes a major burden off them,” Dr. Bell says. Doctors who want to handle anticoagulation themselves may still do so, but the pharmacy insists that they use the same monitoring protocol as the pharmacists.
At Swedish Medical Center in Seattle, the hospitalists are uniformly pleased with a similar setup. When a hospitalist writes an order forwarfarin initiation and establishes a target INR of between two and three, the pharmacists take over.
“That basically gives us the authority to go forward with daily dosing and monitoring,” says Thomas Rivers, PharmD, clinical lead pharmacist at Swedish. It also frees up the hospitalists to focus on other clinical issues.
“We have much more autonomy,” Dr. Rivers says, “and we’re not necessarily having a face to face or verbal conversation with the doctors every day if things are progressing the way we expect. And hospitalists don’t have to be micromanaging these things every day.”