Published in the May 2015 issue of Today’s Hospitalist
TOO OFTEN, doctors have to make care recommendations with only scant or contradictory evidence to guide them.
At this year’s friendly debate about perioperative controversies at the Society of Hospital Medicine’s annual conference, two of the nation’s top perioperative experts squared off on four such topics. Both debaters took turns defending different approaches to those quandaries, laying out various interpretations before sharing what they actually do for real patients in their practice.
And in the end, explained debaters Steven L. Cohn, MD, and Leonard Feldman, MD, hospitalists may need to compromise with anesthesiologists or surgeons who may interpret the same sparse data differently, then ask patients for their preferences before making their recommendations. Dr. Cohn, a professor of clinical medicine at the University of Miami Miller School of Medicine, directs the University of Miami Hospital’s preoperative assessment clinic and medical consultation service. Dr. Feldman is associate professor of internal medicine and pediatrics at Baltimore’s Johns Hopkins School of Medicine and director of the general medicine consultation service at Johns Hopkins Hospital.
Consider the dilemma of aspirin during surgery. “There is no right answer,” Dr. Feldman said. “We don’t have enough evidence.”
In the case presented, a patient with a hip fracture has been taking daily aspirin for secondary prevention after a stroke a year ago, but has stopped taking it for the past three days. Should the hospitalist have the 75-year-old, who is otherwise relatively healthy, restart his aspirin now and take it during surgery? Or should the patient hold off until after surgery when he isn’t bleeding?
During their debate, Drs. Cohn and Feldman both mentioned the same study: the highly publicized POISE-2 trial of aspirin in patients undergoing noncardiac surgery, published in the April, 17, 2014, issue of the New England Journal of Medicine. Study results could convince some hospitalists to not restart the patient’s aspirin until after surgery. But others might reach the opposite conclusion.
Taking the first position, Dr. Feldman noted that about 5% of the patients in the POISE-2 study “were just like our patient” and that the study concluded that “there was no improvement for those patients on aspirin” in terms of preventing death, stroke or heart attack compared to placebo. Moreover, “there was a 1% difference in major bleeding for patients on aspirin.”
One conclusion to draw from that, he said “particularly when considered with findings from another large international study published in the April 15, 2000, issue of The Lancet “is that taking aspirin during surgery “did not help patients in terms of their vascular health, and it did cause more bleeding.”
A key caveat
But as Dr. Cohn explained, the POISE-2 researchers failed to analyze data on an important subgroup: patients with a relatively recent myocardial infarction or stroke who run a higher risk of a perioperative cardiovascular or cerebrovascular event, especially during certain surgeries. This hypothetical patient fell into that subgroup.
“I don’t have a problem with not restarting the aspirin for primary prevention because the risk of a cardiac event is very small,” Dr. Cohn explained. “But I think there is a difference with secondary prevention.”
That’s particularly true because the patient in the case had a relatively recent stroke, as well as hypertension and hyperlipidemia, and he had discontinued his antiplatelet therapy. “These are all things that potentially increase your risk of a postop event, according to the risk calculators.”
In real life, Dr. Feldman said he considers it “reasonable” to recommend either approach “and that involving the patient in a coin-toss decision like this isn’t a bad idea.
As for Dr. Cohn, he said he would probably restart the aspirin, as he argued in the debate, because “I would be concerned about this subgroup” question the POISE-2 trial left unanswered.
That might not be the case for a patient who suffered a stroke 10 years ago. “But with a stroke or MI in the last year or two, I would be concerned,” Dr. Cohn said. “Yes, the patient could bleed, but it’s not likely to be a fatal bleed.”
Another area where hospitalists have “absolutely no evidence,” according to Dr. Feldman, is whether to prescribe supraphysiologic stress-dose steroids for patients who may have adrenal insufficiency.
“It is the standard of care,” Dr. Feldman continued, “but there is very little evidence behind it.”
He and Dr. Cohn considered the case of a 40-yearold woman scheduled to have a total abdominal hysterectomy-bilateral salpingo-oophorectomy in two days. She also has diabetes, hypertension and lupus, and has been taking 30 mg of prednisone daily for the last two months in addition to her blood pressure and diabetes medications. Should the hospitalist during a preop consult recommend high-dose steroids (50-100 mg IV hydrocortisone plus 25-50 mg every eight hours), or just have the patient stick with her current prednisone dose?
When polled, the audience opted overwhelmingly to give high-dose steroids. And the theory behind that strategy makes sense, Dr. Feldman said. Unless the hypothalamic’pituitary’adrenal (HPA) axis is suppressed, the human body can make enough cortisol for both everyday life and for times of stress, “like a big surgery, like getting your uterus taken out.”
The main reason people develop adrenal insufficiency is exogenous use of steroids; adrenal glands atrophy when they aren’t being used. That means, Dr. Feldman explained, that “when you need your adrenal glands to work in time of stress, they are not there for you.” Because the patient has been taking prednisone for more than three weeks, she will “at least theoretically “be at high risk for adrenal insufficiency during moderately stressful surgery. The consensus, said Dr. Feldman, is that physicians should order the IV steroids for her just before surgery and taper them down afterwards.
But this is all theoretical, Dr. Cohn argued, and based on several extremely small and mostly descriptive or retrospective studies. And some of those studies have found a poor relationship between adrenal insufficiency and perioperative hypotension.
Only two randomized controlled trials, which looked at a total of 37 patients, delved into this topic, and those were inconclusive. It may well be, Dr. Cohn concluded, that patients whose HPA axis is suppressed “can get by on their usual amount,” even during stressful surgery.
In reality, Dr. Cohn said that he regularly gives stress-dose steroids to patients, “even though I know there is no evidence.” Dr. Feldman pointed out that a few studies currently underway should give physicians more information to work with. Until then, he said, “I think we have to stick with the standard of care.”
ACE inhibitors and ARBs
One area where there are data “but “very conflicting” data “is the perioperative use of ACE inhibitors and ARBs, Dr. Cohn said. How should you balance the risks of hypotension and worsening renal function associated with anesthesia against the cardioprotection and hemodynamic-stability benefits of keeping patients on their ACE inhibitors and ARBs?
The case debated concerned a 65-year-old woman with hypertension, coronary artery disease, a five-year-old stent and diabetes (all pretty well-controlled by medications) about to undergo colon-cancer surgery. Should you hold her ACE inhibitor on the morning of surgery and restart it when she is hemodynamically stable, or keep her on the drug throughout the surgery?
Anesthesiologists worry that patients taking ACE inhibitors or ARBs might suffer hypotension during surgery, said Dr. Cohn. But the studies that have looked specifically at that risk “are all over the board,” he explained. “Some show hypotension, some don’t.”
Moreover, Dr. Cohn explained that hypotension is “usually easily treated with no evidence of poor outcome.” Complicating matters are the different ways practitioners and researchers define hypotension and the differences among specific agents. “It may be that ARBs are worse than ACE inhibitors and that enalapril is worse than captopril,” he said. The drugs’ effect may also be exacerbated if patients are on a diuretic.
And in terms of whether blood pressure drugs can affect the kidneys, Dr. Cohn added, the results are much the same: “mixed.”
On the plus side, some limited studies have found that these drugs offer cardioprotection and that withdrawing them can make chronic heart failure and myocardial ischemia worse. An important meta-analysis published online in August 2008 by the Journal of Hospital Medicine recommends that “you withhold them to prevent hypotension but continue them to prevent MIs,” Dr. Cohn said. “How are you supposed to do both?”
Will you remember to restart?
Even more concerning: Once a drug is stopped, it may never be restarted. Nonresumption of ACE inhibitors within 14 days of surgery is associated with increased 30-day mortality.
“My concern is that if you stop the drug but don’t restart it in the hospital,” Dr. Cohn said, “the patient will go home without it and be at increased risk for a postop cardiac event.”
While ACE inhibitors can help lower length of stay and the risk of perioperative MI and mortality, Dr. Feldman noted that the side effect of vasoplegia shouldn’t be discounted. It can be a “really big problem” not only for anesthesiologists but for patients who suffer acute kidney injury as a result. Plus, many patients take ACE inhibitors and ARBs for reasons other than heart failure or coronary artery disease.
“Does it matter if you are holding it in your heart failure patients vs. those with hypertension?” he asked. “It’s not quite clear, and we may get different results with different surgeries.”
Dr. Feldman makes the decision to hold or not hold on a case-by-case basis. “It depends on the patient’s indications for getting that ACE inhibitor,” he added. “If the patient has essentially normal blood pressure but is on it for proteinuria, I will hold it. In a heart failure patient who’s already hypotensive, I don’t know that it makes sense to hold it.” But if you do stop a medication, he added, “make sure you restart it so patients go home on their ACE inhibitor.”
Dr. Cohn agreed, adding that part of the decision has to be made in conjunction with the anesthesiologists. And sometimes, you’ll need to compromise after looking at the patient’s blood pressure and heart failure status. Another strategy, he added, is to give the drug the night before, not the morning of, surgery.
HbA1c level cutoffs
By a margin of nearly three to one, audience members said they wouldn’t recommend canceling elective surgery just because a patient’s glycosylated hemoglobin level exceeded a certain percentage, such as 7, 8 or 9. Nonetheless, said Dr. Feldman, hospitalists should be aware that some orthopedic surgeons (and hospitals) are setting thresholds above which they will not operate, so hospitalists will have to discuss this with their surgeons.
Evidence confirms that having diabetes puts surgical patients at increased risk for many complications, from death to infections. But it isn’t clear what the HbA1c level has to do with that.
“There is poor and conflicting evidence as to what an A1c predicts as far as outcomes,” Dr. Feldman explained. Moreover, at what level does the complication rate become unacceptable?
“If we target a too high threshold,” he said, “too many complications will take place. If it’s too low, we’ll be delaying or denying surgery to patients who would otherwise not have complications. There has to be a right number that we can pick to make sure enough people have access to surgery but that will help us convince those who are going to have complications not to have surgery.”
The case they considered was a 62-year-old woman scheduled for elective hip replacement in two weeks who has refused to take insulin for her uncontrolled diabetes. Her HbA1c is 9.5, even though she takes metformin and glipizide at home.
“Is this a teachable moment to motivate patients?” Dr. Feldman asked. One could argue that this patient should be told she can’t get hip surgery unless she reduces her HbA1c.
Although there are no randomized controlled trials, some observational data “particularly from a study in the March 19, 2014, issue of the Journal of Bone & Joint Surgery “suggest that threatening loss of surgery might work. But in that study, HbA1c reduction took many months (the median was between four and five), and some patients couldn’t reach a 7% cutoff at all.
One problem with relying on HbA1c levels to make major decisions about surgery, said Dr. Cohn, is that it is “a poor marker for complications.” Instead, hyperglycemia “either preop or postop “seems to be more important. A VA study published online in January 2015 by The Journal of Arthroplasty concluded that “while HbA1c did not perfectly correlate with the risk of periprosthetic joint infection, perioperative hyperglycemia did.”
Dr. Cohn also presented evidence that low HbA1c levels (below 6.5% or 6.7%) were associated with increased lengths of stay or wound complications. “In my opinion, there are too many variables, so I don’t think we can say there is an A1c that should be a cutoff,” he said. “This may be something we do for no reason.”
In their actual practice, both Drs. Cohn and Feldman expressed concern about setting “arbitrary cutoffs.” But they both agreed that hospitalists doing preop assessments in conjunction with orthopedic surgeons should use the “teachable moment.”
“You may get some patients to start taking their diabetes seriously if you tell them they have a higher risk for certain surgical complications if they don’t get their A1c level and their diabetes under control,” Dr. Feldman said. “But I would not deny access to patients if they are unable to reduce it.”
Deborah Gesensway is a freelance writer who covers U.S. health care from Toronto.