Published in the April 2010 issue of Today’s Hospitalist
It’s a busy Friday morning when you encounter Ms. Smith, a 66-year-old patient with a prior history of gastric ulcers admitted the previous evening for upper abdominal pain and scheduled to undergo an upper endoscopy. The primary care physician, who wants you to manage the patient, tells you that the patient has been cleared by the gastroenterologist, who may do a biopsy.
What the outpatient physician fails to mention to the GI, however, is that Ms. Smith has chronic atrial fibrillation, a past history of cardioembolic stroke and is on warfarin.
When she clocks in with an INR of 2.9, the GI refuses to proceed unless the patient’s INR is "substantially reduced," preferably to just over 1.0. That leaves you stuck in the middle of a political and clinical dilemma. Should you try to get the patient’s INR down quickly with a transfusion of fresh frozen plasma (FFP)? Or should you allow the INR to "drift down," which may require bridging and a longer, more costly inpatient stay?
While this standoff is common, even experts admit that there’s often no easy answer. "Sometimes I feel we spend half our lives as hospitalists balancing bleeding and clotting risk," says David Feinbloom, MD, an academic hospitalist at Beth Israel Deaconess Medical Center (BIDMC) in Boston. "This is a constant source of confusion and negotiation because we deal with so many patients on anticoagulants."
For Dr. Feinbloom, assistant professor of medicine at Harvard Medical School, part of the solution has been to develop periprocedural guidelines for managing patients on long-term anticoagulation or antiplatelet therapy.
While evidence is slim, Dr. Feinbloom suggests that hospitalists sit down with proceduralists to hash out guidelines they can at least agree on “keeping in mind, he says, that there will always be scenarios that will need to be addressed on a case-by-case basis.
Conflicting specialty recommendations
Alan Brush, MD, chief of the anticoagulation management service for Harvard Vanguard Medical Associates in Boston and assistant clinical professor of medicine at Harvard Medical School, is well aware of how many times hospitalists hit this roadblock. He receives a half-dozen calls a day regarding target INR and bridging decisions.
Many times, Dr. Brush is called to intercede because of differences in specialty recommendations. "The hospitalist might say, ‘I want this patient anticoagulated with an INR of 2.0 to 3.0,’ " he says. "But the surgeon might cite American College of Orthopedic Surgeons recommendations calling for an INR of up to 2.0 because of bleeding risk."
Things get even thornier with individual provider preferences. Although the American Society for Gastrointestinal Endoscopy late last year developed a list of procedures that can be performed safely for patients taking warfarin, for example, some proceduralists just ignore it.
"When I surveyed all of our gastroenterologists, none would work on patients on anticoagulation if they were doing a biopsy," Dr. Brush says. "If no one will do the procedure, the hospitalist must decide if the patient requires a bridge and, if so, how best to bridge the patient."
What often happens, Dr. Feinbloom observes, is that time is wasted delaying procedures and arguing about what the INR value should be. "There are proceduralists who won’t touch a patient unless the INR is 1," Dr. Feinbloom says. "But even large FFP transfusions may not completely correct mild to moderate INR elevations." While no data suggest that prophylactic periprocedural plasma transfusion prevents serious bleeding complications, he adds, a unit of donated plasma may have an INR of up to 1.3. That makes it not only unnecessary but "very difficult," Dr. Feinbloom says, to lower an INR below 1.5 with plasma transfusion.
When it comes to target INRs, anticoagulation experts often opt for a solution that Dr. Brush calls a "practice compromise." For patients with moderate bleeding risk, for instance, guidelines recommend a target INR of 1.5-1.7 the day of the procedure, meaning that warfarin must be held. When bleeding risk is high, however, or a procedure calls for epidural anesthesia or steroid injection, an INR of 1.3 or less is needed.
Many anticoagulation experts would rather entertain bleeding risk than stroke risk “except for surgical or diagnostic procedures in which even a small bleed could be catastrophic. That includes most neurosurgeries or procedures in an organ or system, such as the epidural cavity, that can’t be tamponaded. Unfortunately, experts say, data consist mostly of retrospective, not controlled prospective studies.
At Beth Israel Deaconess, Dr. Feinbloom and his colleagues have tried to synthesize specialty-society guidelines and supplement them when possible. They then literally "shopped" their initial guidelines on target INRs and bridging among BIDMC’s proceduralists. The procedural division chiefs signed off on the modified document.
The guidelines say, for instance, that most elective endoscopic procedures that don’t involve biopsy can be done safely without stopping warfarin or antiplatelet agents because the bleeding risk is relatively low.
But the document calls for holding warfarin until the INR is 1.5 or less if biopsies or polypectomies are involved, and bridging with unfractionated heparin (UFH) or low molecular weight heparin (LMWH), based on individual patients’ risk for thromboembolic events. While individual proceduralists may still decide not to perform a particular procedure on anticoagulated patients, hospitalists "can at least say, ‘Well, here’s a multidisciplinary guideline that your chief signed off on that says it’s safe,’ " Dr. Feinbloom says.
Where to start
Many institutions base their guidelines largely on those issued by the American College of Chest Physicians (ACCP) and published in the June 2008 issue of CHEST. Another set often used is the American College of Cardiology/ American Heart Association guidelines issued in the March 2002 Journal of the American College of Cardiology on both general and perioperative management of atrial fibrillation.
The CHEST guideline calls for stratifying atrial fibrillation patients’ peri- and post-op stroke risk using clinical prediction rules. The most widely accepted is the six-point CHADS-2 (Congestive Heart Failure-Hypertension- Age-Diabetes-Stroke) scoring index.
That index targets patients with nonvalvular atrial fibrillation, heart failure, hypertension, age 75 or older, and diabetes, all of which receive one point. The same index, however, gives two points each for prior stroke, TIA or peripheral embolism. Generally, stroke risk rises with progressively higher CHADS-2 scores, and patients at the higher end may benefit more from minimizing their time off anticoagulants.
"CHADS-2 isn’t great, but it’s the best we’ve got," says Joseph Dorsey, MD, a Walpole, Mass., physician and medical consultant to the Massachusetts Coalition for the Prevention of Medical Errors on strategies for reducing the risk of long-term anticoagulation. He formerly directed Harvard Vanguard’s hospitalist program
The 2008 CHEST guidelines call for simply holding warfarin in low-risk patients or procedures, but advise bridging with therapeutic dose UFH or LMWH in high-risk patients with atrial fibrillation, VTE or mechanical heart valves.
Even when institutions use CHADS-2 to guide risk stratification and bridging, physicians may tweak the guidelines. "Our guidelines are more conservative than CHEST’s," says Dr. Feinbloom. "We have a lower threshold to bridge than the ACCP recommends and are more conservative on the use of LMWH."
Bridging: full or preventive dose?
But while UFH continues to be the standard, groups are experimenting with LMWH to keep patients being bridged from being admitted for monitoring.
"The long-term traditional way to bridge has been IV-monitored heparin," says Dr. Brush. "But whenever this has been compared with LMWH, LMWH almost always ends up comparing better or at least as well as IV, and it’s a lot easier to manage." The exceptions are patients with severe renal insufficiency or extreme obesity where subcutaneous LMWH can be both risky and hard to manage.
He recommends LMWH dosing (with Lovenox) at 30 mg twice daily as a prophylactic dose, and either 1 mg per kilogram twice daily or 1.5 mg once daily as a full dose.
"One decision to be made with the proceduralist when the patient resumes LMWH is this: Is it safe to resume at the full treatment dose or better to use a preventive dose?" says Dr. Brush. While the lower dose may not give you full protection, "it gives some protection and is much less likely to cause bleeding. It gets complicated, and it requires frequent conversations."
Dr. Brush points out, however, that bridging decisions for a host of procedures are slam dunks. Those include patients with atrial fibrillation who have had a relatively recent cardioembolic stroke, mechanical mitral valves or combinations of multiple risk factors for stroke and/ or VTE.
"These are the patients who make you the most nervous, so the plan is to bridge unless you can’t," he says. "You definitely want to limit the period when patients are below therapeutic range as much as possible, except that you don’t want to cause a new bleed." He recommends refraining from restarting LMWH until you can reasonably assume that bleeding is not occurring.
Still, the guidelines don’t even begin to address all the potential nuances of bridging decisions, says hospitalist Margaret Fang, MD, medical director of the University of California, San Francisco (UCSF) anticoagulation clinic. One big problem is that there isn’t any randomized trial evidence that hospitalists can use.
However, one current 40-center trial “the BRIDGE trial, which began last year and will be completed in 2013 “involving more than 3,600 patients with atrial fibrillation may "give us more evidence regarding the comparative risks of holding perioperative bridging vs. providing it," says Dr. Fang.
Making patients part of the decision
In the meantime, Dr. Fang urges hospitalists to take another set of preferences into account: the patients’. She uses the CHEST guidelines as a starting point not only for discussions with proceduralists, but with patients "about the risks of either remaining on full-dose anticoagulation or half dose, or nothing."
For many anticoagulated patients, the risk of first stroke or recurrent stroke may be paramount, regardless of what guidelines say about the safety of certain procedures with or without bridging.
She cites a recent 68-year-old patient on chronic anticoagulation for atrial fibrillation. A GI was urging the patient, who had a mechanical heart valve, to undergo a screening endoscopy that would require bridging. When the patient understood how complicated the procedure could become from the standpoint of stopping warfarin and being bridged with a second medication, she decided to forgo the procedure.
"She basically said, ‘I’m willing to have something missed on a screening endoscopy,’ " Dr. Fang reports. " ‘I’d rather leave things the way they are and not go off the anticoagulant.’ "
The whole issue of perioperative anticoagulation may soon be less complicated, Dr. Dorsey predicts, when one or more of the new class of direct thrombin inhibitors receives FDA approval. One huge advantage of those anticoagulants, including dabigatran, is that the INR does not need to be monitored.
"These experimental drugs will likely take the place of warfarin “and when they do, life will get a lot easier," Dr. Dorsey says. "But we’re not there yet."
Bonnie Darves is a freelance health care writer based in Seattle.
Dispelling management myths
AMIR JAFFER, MD, A NATIONAL EXPERT in perioperative medicine and atrial fibrillation, recommends a practical approach to bridging decisions. Dr. Jaffer, the division chief of hospital medicine at the University of Miami, is co-author of the 2008 CHEST guidelines and is on the panel for the 2012 version of those guidelines as well. He notes that several myths persist about perioperative anticoagulation management.
The first myth is that warfarin must be discontinued for all procedures, which is not true. "With the major procedures you must stop warfarin, but with the minor ones it needn’t be discontinued, based on systematic reviews," Dr. Jaffer says. Major procedures include those with a relatively higher bleeding risk or high risk if any bleeding occurs (such as neurological procedures) or a protracted inpatient stay for individual patients.
According to Dr. Jaffer, the jury is still out on when bridging is indicated in atrial fibrillation patients. That question won’t be decided, he notes, until the randomized BRIDGE trial, which is enrolling more than 3,600 patients, is complete. For major surgery, he recommends discontinuing warfarin five days before the procedure if the peri-op INR target is between 2 and 3, six days if it’s between 3 and 4.5. The goal, he says, "is to get the INR to less than 1.2 before the procedure." Low molecular weight heparin (LMWH) at a dose of 1 mg/kg or dalteparin at 100 IU/kg should be started 36 hours after the last warfarin dose.
Although the dosing issue remains a difficult one, Dr. Jaffer points out that the full-dose vs. prophylactic heparin question is unresolved. A study in the February 2010 American Journal of Medicine, however, found that bridged patients who received full-dose unfractionated heparin or LMWH had a six-fold postop bleeding rate compared to those receiving lower doses or no bridging. That finding, he says, seems to endorse less aggressive bridging post-op, especially during the first 48 hours after major surgery.
"The centers that used heparin bridging more aggressively, especially in the post-operative period, experienced higher rates of bleeding," says Dr. Jaffer. "Center strategy turned out to be the most important predictor of bleeding rates."
CHECK OUT THE CLINICAL PROTOCOL WEB PAGE for the Harvard Vanguard Medical Associates’ anticoagulation management services guideline. Go online to the Clinical Protocols link at www.todayshospitalist.com.