Published in the June 2014 issue of Today’s Hospitalist
SOMETIMES IN PERIOPERATIVE MEDICINE, there is no clear right or wrong approach. Whether it’s choosing a beta-blocker before vascular surgery or deciding whether to order an echocardiogram before hip-fracture repair, there are times when smart hospitalists can agree to disagree.
In this vein, two leading consultative medicine experts debated four controversial perioperative topics at this spring’s Society of Hospital Medicine’s annual conference. Stressing that they were arguing both sides of the issues to get hospitalists thinking, panelists Steven L. Cohn, MD, director of the University of Miami Hospital’s preoperative assessment clinic and medical consultation service, and Kurt Pfeifer, MD, director of the preoperative evaluation clinic at the Froedert and Medical College of Wisconsin Eye Institute in Milwaukee, explained that they wouldn’t necessarily choose the approach they were advocating for in the debate. Instead, they were laying out valid “but differing “interpretations of complicated topics, often based on the same evidence.
Beta-blockers and vascular surgery
Consider this question: Should you recommend changing a 70-year-old man’s beta-blocker drug one week before he is scheduled for femoral-popliteal bypass surgery? In the case the debaters described, the patient had been taking short-acting metoprolol. Most of the evidence garnered from a decade of research, however, shows that more cardioselective beta-blocker medications “particularly atenolol and bisoprolol “tend to be associated with lower cardiac, stroke and mortality risk than metoprolol when used perioperatively, Dr. Cohn explained.
The issue is further complicated by evidence that postoperative outcomes are affected not only by the choice of beta-blocker, but also by how long before surgery patients start taking it and whether the dose is titrated adequately to achieve heart rate control. Studies have shown an association between patients starting beta-blockers more than a week before surgery and both a lower preoperative heart rate and improved outcomes, as compared to those starting less than a week before surgery.
When “a high-risk patient is going for high-risk surgery in seven days,” Dr. Cohn reasoned, the evidence argues that there is adequate time to switch the patient to a more cardioselective drug. “I would switch this patient from metoprolol and titrate the dose in the seven days that I have before surgery, rather than leaving him on the drug that is going to be less effective.”
In real life “outside a debate forum “however, Dr. Cohn, told the audience that he was more likely to do as Dr. Pfeifer argued: not “switch someone in midstream” with just a week to go before high-risk surgery.
If there was time to “start from scratch well in advance” of surgery, Dr. Pfeifer said, he would interpret the evidence as advising hospitalists to “try one of these once-a-day, cardioselective beta-blockers. But if you are in a high-stakes situation, and you already have ‘a good hand’ ” “meaning the patient is well-controlled on a decent drug “”I would stay with what I’ve got because you are much more likely to get a benefit from that than from switching therapy. You don’t want to let the perfect be the enemy of the good.”
Aortic stenosis & hip fracture
Another topic debated was what a hospitalist should do when called to evaluate a patient with a systolic murmur compatible with aortic stenosis just before she is slated for orthopedic surgery to repair a hip fracture. Should you order echocardiography?
Dr. Pfeifer said he “would definitely get an echo,” particularly because it can be done quickly and not delay urgent surgery. Information gleaned from the ultrasound scan would help both the anesthesiologist during surgery and the hospitalist providing postoperative care.
It is also important to get the diagnosis right, he added, because aortic stenosis complicates surgery. Some studies have shown that when aortic stenosis is severe, perioperative mortality is about 10% and patients have increased risks of MI, heart failure and stroke. In addition, he said, hip fracture surgery is “urgent” but not “emergent,” so doctors still have some time to get the echo.
What would he do with the findings? Dr. Pfeifer said he would not delay the hip fracture surgery by recommending valve surgery. Instead, he would pass the information about the severity of any stenosis to the anesthesiologist, who could then determine the best way to sedate and monitor the patient.
The data would also help with postop triage. “In a situation with a large amount of blood loss and large changes in blood pressure, I would want the anesthesiologist to know if the patient had aortic stenosis so he or she could determine the best postoperative management setting,” Dr. Pfeifer said.
On the flip side of the debate, Dr. Cohn argued that hospitalists shouldn’t underestimate how important it is to repair hip fractures quickly. A patient with severe aortic stenosis would not be a candidate for aortic valve replacement, he noted, and sending the patient for an echo in itself would delay surgery. Meanwhile, the condition can often be clinically diagnosed, and doctors can have a frank conversation with the anesthesiologists even without echo results in hand. Getting the echo, he said, probably won’t change management.
“Patients who have significant aortic stenosis are at increased risk for complications, but there is no way that getting a preop echo is going to decrease that risk,” Dr. Cohn said. “The bottom line is that asymptomatic AS and severe AS patients who are not candidates for valve replacement can undergo surgery at an acceptable risk. Patients with symptomatic severe aortic stenosis should have elective noncardiac surgery delayed, but hip fracture surgery is not elective.”
In this controversy, the two presenters actually disagreed in real life. Dr. Cohn told the audience he would probably opt to do as he had argued: “Just tell the patients they probably have AS and have the anesthesiologist treat them accordingly and not necessarily get an echo. I think it’s an extra cost” that wouldn’t change how patients were treated and managed.
Dr. Pfeifer, on the other hand, would “definitely get an echo” for this patient, “and I would have it done quickly, and I would have significantly improved the patient’s perioperative care by knowing exactly how we need to manage her.” He referred to literature that shows that clinical diagnosis of aortic stenosis is unreliable. Assuming a patient has aortic stenosis and providing empiric intraoperative management could be harmful to a patient without the echo.
Anticoagulation and neurosurgery
In the case of a 50-year-old woman who develops a pulmonary embolism four days after a craniotomy to remove a meningioma, should a hospitalist consultant recommend full-dose anticoagulation or an inferior vena cava (IVC) filter?
The evidence shows many tradeoffs between being on anticoagulants and chancing a horrendous intracranial hemorrhage, on the one hand, vs. implanting a less effective IVC filter, even one that is retrievable. The filter, after all, may not prevent recurring venous thromboembolism (VTE) as well as it should, and filter use has been associated with complications.
According to Dr. Cohn, a number of studies show that the bleeding risks of anticoagulation may be less than people fear and that “filter complication rates” are “higher than expected and may outweigh risks of anticoagulation, particularly because temporary filters are not removed as often as they should be.”
He described a study in the January 2014 issue of the Journal of Neurological Surgery, which looked at 42 patients with VTEs after craniotomy. The authors concluded that “more aggressive management regarding anticoagulation” with enoxaparin or unfractionated heparin “may be justified” and that the medical profession perhaps should be “getting away from being afraid to use” anticoagulation in these patients.
“The problem with filters is that there is limited evidence for what they do,” Dr. Cohn said. “They do prevent VTEs, but there is no reduction in mortality, and a higher incidence of long-term DVT” is associated with their use.
He argued that hospitalists should discuss the pros and cons of both strategies with the surgeon. If the surgeon feels there is no significant postop bleeding or oozing, Dr. Cohn said, “it would be reasonably safe to treat a patient with VTE with therapeutic anticoagulation.”
Dr. Pfeifer championed the other side, saying that anticoagulation and intracranial surgery are a dangerous combination and that “even a small amount of hemorrhage can be devastating.”
Even if evidence shows benefits from more aggressive anticoagulation, he added, “there are so many variables” that can cause bleeding in neurosurgery patients that it makes it very difficult to broadly recommend anticoagulation. “We don’t know if some of those other things might cause severe bleeding.”
Despite the limitations of IVC filters, Dr. Pfeifer explained, this patient has “the one true evidence-based indication for IVC filters. I just want something temporary to keep the patient from having a problem with bleeding from anticoagulation or recurrent clot.”
As for retrievable filters not being taken out as they should, Dr. Pfeifer said this is a different problem that should be addressed by always having a plan for removal when the filter is placed. “It doesn’t make any sense to not use something that is appropriate just because some people use it wrong,” he explained.
What would happen in the real world? Both Drs. Cohn and Pfeifer would do the same thing for this patient: Discuss the risks of intracranial hemorrhage with the neurosurgeon and work with him or her on the decision for a filter or anticoagulation. “I don’t think either of us would push for anticoagulation if the neurosurgeon expressed serious concern for bleeding,” Dr. Pfeifer said.
Atrial fibrillation and bridging
When a 70-year-old man with A-fib undergoing total knee replacement needs to temporarily stop his warfarin, should a hospitalist recommend perioperative anticoagulation bridging therapy when the patient’s CHADS2 score is 3?
According to Dr. Cohn, he “wouldn’t bridge” this patient. The evidence to date, although not strong, shows that “the risk of thromboembolism without bridging is low” for this type of patient while “the risk of bleeding with early postop anticoagulation is significant.”
A study published online in 2010 in the Journal of Thrombosis and Haemostasis also found that bridging didn’t reduce stroke risk in patients with A-fib.
“There was no reduction in thromboembolic events with bridging vs. without bridging because these events almost never occur,” said Dr. Cohn.
He also cited a 2012 systematic review in Circulation of all the randomized controlled trials on the topic to date. It concluded that perioperative heparin bridging was associated with an increased risk of overall and major bleeding, with the same risk of thromboembolic events compared to comparable patients who weren’t bridged.
Dr. Cohn also pointed to a review article in the May 30, 2013, issue of the New England Journal of Medicine, which recommended that patients with a CHADS2 score of 3 or less do not need bridging.
“The premise that bridging anticoagulation should be the default mechanism until we show that it’s effective is an example of well-intentioned but misguided treatment,” Dr. Cohn said. “I think we are too afraid of things and we may be causing more harm, so I say don’t bridge.”
Dr. Pfeifer, however, argued the pro-bridging side. He reminded hospitalists that the evidence against bridging isn’t strong “and that the automatic conclusion that these patients shouldn’t be bridged may be premature. Two large national studies on perioperative bridging are underway, he pointed out, “so we will hopefully have an answer” in the next few years.
Although atrial fibrillation is a strong predictor of morbidity and mortality, guidelines published by the American College of Chest Physicians in 2012 say that patients with CHADS2 scores of 5 and 6 should be bridged perioperatively, while those with scores of 0-2 should not. For those scoring 3 and 4, “it’s all based on your discussion of values with the patient,” Dr. Pfeifer said.
But when discussing the bridging option with patients, Dr. Pfeifer said he worries that the numbers sometimes don’t tell the whole story.
“It’s important to talk about the consequences of a bad outcome, not just the numbers,” he said. “For arterial thromboembolism, 20% are fatal, and more than 50% result in permanent disability.”
With major bleeding events, on the other hand, “about 10% are fatal and hardly any of them result in permanent disability. So yes, even if bleeding happens more often than stroke, stroke is what is really going to take you down and destroy your life. That is what I am worried about.”
Like most of the audience members, Dr. Cohn said he would not bridge a patient of his in this situation. And in this case, Dr. Pfeifer actually agreed.
Deborah Gesensway is a freelance writer who covers U.S. health care from Toronto.