Published in the October 2017 issue of Today’s Hospitalist
A 71-YEAR-OLD MAN admitted with cellulitis has a history of hypertension. He’s taking two antihypertensives, and he had an MI with PCI three years before. On hospital day 2, his blood pressure is elevated, although he’s having no symptoms and you don’t suspect end-organ damage.
Should you administer a one-time antihypertensive dose if that elevated pressure meets the definition of hypertensive urgency, with a systolic of 180 mmHg or more and a diastolic of 110 mmHg? If not at 180/110, how about at 205/110? Or 215/119?
“As the numbers creep up, we start to get nervous, and we think that something bad is going to happen over the next few hours if we don’t treat this,” said hospitalist Anthony C. Breu, MD, director of medical resident education at the VA Boston Healthcare System in West Roxbury, Mass. Those “bad” things include acute stroke, MI or aortic dissection. Dr. Breu spoke about treating hypertensive urgency as part of this year’s “things we do for no reason” presentation at the Society of Hospital Medicine annual meeting in Las Vegas.
“Even with patients who have severe hypertension, it takes time for bad things to happen.”
“The current standard of care suggests we ought to treat, and we’ll keep getting paged until we do,” he explained. “Many of the people we consult with and nursing staff see us as being lazy or inattentive if we don’t treat.” A study in the March 2016 issue of the Journal of Hospital Medicine, for example, found that more than 98% of episodic IV antihypertensive orders in one academic center were for systolic blood pressures under 200 mmHg, while 84.5% were for systolic pressures under 180 mmHg.
But Dr. Breu said that evidence has raised a lot of questions about whether treating hypertensive urgency is the right thing to do. To make his case, he cited several studies that, taken together, indicate that treating hypertensive urgency may not only be unnecessary, but could be harmful.
A chronic and cumulative disease
The first study Dr. Breu discussed was 50 years old, published in the Dec. 11, 1967, issue of the Journal of the American Medical Association. But that research it’s still important, he said.
Male patients with hypertension whose diastolic blood pressures averaged between 115 and 129 mmHg were randomized to either active treatment (hydrochlorothiazide plus reserpine plus hydralazine hydrochloride) or placebo. “Epidemiology in the 50s and 60s suggested that diastolic hypertension was a more important risk factor than systolic,” he pointed out. While the authors reported on eight different outcomes, “if you were in the treatment arm, you did pretty well.” The number needed to treat to prevent death, stroke, heart failure or MI was 5.
“With treatment, risk fell from 40% to 3%,” Dr. Breu pointed out. “In internal medicine, we don’t see absolute risk reductions like that anymore.”
Why was he talking up this study to make the case for thinking twice about treating hypertensive urgency? Because in the treatment arm, the average time to patients’ first event was one month and two months for those in the placebo arm. “Hypertension is a cumulative chronic disease,” he said. “Even with patients who have severe hypertension, it takes time for bad things to happen.”
In another study published in the July 1, 2016, issue of JAMA Internal Medicine, authors looked at what happened to patients who had hypertensive urgency in an outpatient clinic. “They were particularly interested in how often patients had acute coronary syndrome, a stroke or a TIA,” said Dr. Breu. Authors compared outcomes for patients sent to a hospital vs. those sent home.
Out of nearly 60,000 patients, more than 99% were sent home and only 100 were admitted to a hospital. “Most did fine in both arms,” Dr. Breu said, “and even at seven days, the event rate was exceedingly low: only 1 in 1,000 patients.”
What about major adverse cardiovascular events in patients with more than 220 systolic? “Same story: Two patients in 1,000 had an event within seven days,” Dr. Breu said. “The key point is that these events are exceedingly rare.”
In a third study, which appeared in the September 2008 issue of The Journal of Clinical Hypertension, nearly 550 patients with hypertensive emergency were seen in the ED. Patients were first directed to rest for 30 minutes.
“In half an hour, the blood pressure of one-third of them dropped below the urgent range,” he said. Those whose blood pressure didn’t fall that far were then randomized to different medications (perindopril, amlodipine or labetalol), then had their blood pressure reassessed at 60- and 120-minute intervals.
Unfortunately, said Dr. Breu, “one in six patients had a blood pressure drop of more than 20% within the first few hours, which you don’t want.” That included close to one-third (31.4%) of those receiving labetalol— “and I’m sure there are many people in the room who are big labetalol prescribers,” he told the audience.
“Don’t treat a number”
Based on those results, Dr. Breu argued that treating hypertensive urgency could actually be harmful. “What might happen if cerebral blood flow begins to fall off?” he asked. “Strokes can happen, and there are probably people here who have either administered or ordered sublingual nifedipine to treat hypertensive urgency and saw strokes or heard of strokes because of a precipitous drop in blood pressure.”
His recommendation: “Don’t treat a number, even if it gets to 220 or more,” said Dr. Breu. “Really evaluate patients, make sure they don’t have evidence of end-organ damage, let them rest, and then come back and recheck. The number itself shouldn’t be the barometer for administering antihypertensives.”
And while many people may disagree with him, “we should take advantage of hospitalization to potentially tweak outpatient regimens,” Dr. Breu said.
That’s particularly true for patients who are chronically hypertensive in clinic visits as well as during hospitalization, with “changes made ideally in concert with patients’ outpatient physicians.”
Meanwhile, hospitalists should work to change the standard of care in their facilities around treating hypertensive urgency. A good place to start, he suggested, is getting rid of the term altogether. He pointed out that the JNC-7 recommendations issued in 2003 clearly state that the use of “hypertensive urgency” has promoted treatment, while the JNC-8 released in 2014 “doesn’t use the term at all.”
Unnecessary blood cultures
Dr. Breu then turned to another common practice backed by no evidence: ordering blood cultures for inpatients already on antibiotics. “We do it, even when patients have prior negative cultures.”
He understands why doctors have a low threshold for ordering. “While mortality for bloodstream infections has gone down in recent decades, it’s still more than 10%,” he noted. “Blood cultures remain the gold standard for diagnosing them.”
At the same time, the Infectious Diseases Society of America “doesn’t give us much guidance on when to culture, so we’re left with our habits and what we learned where we trained.” And “culture if spikes” remains one of the most commonly used anticipatory orders in written signouts, as shown in a study in the August 2009 issue of BMJ Quality & Safety.
To reduce orders for repeat cultures, consider this question: How often do doctors discover a true positive or a new pathogen?
For answers, Dr. Breu again cited three studies, including one published in the June 2001 issue of Clinical Infectious Diseases. All 139 patients in this single-site study were taking antibiotics. The authors found not a single new positive among the repeat cultures of patients whose previous cultures were negative.
For patients with previously positive cultures as well as positive repeat cultures, “it was almost assuredly what patients were growing before,” he pointed out. “If you were staph before, you’re staph now.” Only one patient with a previously positive culture was found to have a new pathogen on repeat culture.
In the next study, published in the July 2004 issue of Clinical Microbiology and Infection, patients with repeat cultures on antibiotics had very similar outcomes: Only three patients out of 100 grew a new organism on antibiotics.
And Dr. Breu was co-author of the third study, which was published in the May 2016 Journal of Hospital Medicine and looked at the indication for— and the yield of—blood cultures ordered on a medical service. In that study, only one new pathogen was found in a patient taking antibiotics.
“If we combine these three studies,” he noted, “the true positive rate for a new pathogen among patients already on antibiotics was less than 1%.”
Then there’s the issue of contaminants with blood cultures: Studies have found that false positives due to contaminants produce longer lengths of stay with higher associated costs. One study, published in the Journal of Hospital Infection in March 2011, found that false positives from contaminated blood cultures over one year in one teaching hospital added nearly 1,400 unnecessary hospital days and nearly $2 million in costs.
Wait for results
Dr. Breu offered two recommendations. First, for doctors who still aren’t convinced to cut back on orders, studies suggest that culturing patients when their antibacterial activity in the blood is at its lowest increases the yield of a true positive. “So order blood cultures at the trough level of the antibiotic,” he said.
“That increases the probability of a true positive.” However, he added, “we don’t check troughs with most antibiotics, so if you’re not going to order at the trough, just add the culture to the patient’s morning labs,” instead of ordering “culture if spikes.”
“Patients are already being stuck for morning labs,” Dr. Breu said, “and you’re just as likely to get a true positive in those labs as you are with a 2 a.m. fever.”
And for patients on antibiotics for whom the ED or the admitting team has already sent cultures, “don’t order more until those cultures have results,” he advised. “Then base your decision on whether to order more on the pathogen found.” For patients positive for S. aureus, for instance, “you should get surveillance cultures. But with gram negatives or strep, most guidelines suggest you don’t need those unless patients’ clinical situation changes and they become septic and hypotensive.”
And for patients whose initial cultures come back negative? “Ordering more is really unlikely to find anything relevant.”
Phyllis Maguire is Executive Editor of Today’s Hospitalist.
What hospitalists can learn from ED docs
IN ARGUING to not reflexively treat hypertensive urgency, Anthony C. Breu, MD, a hospitalist and the director of medical resident education at the VA Boston Healthcare System in West Roxbury, Mass., pointed out that the emergency medicine community “is ahead of us on this.”
Some studies, for instance, indicate that more than 80% of hospitalists would order IV antihypertensives for patients with blood pressures under 180 mmHg. But survey results published in the March 2017 issue of The Journal of Clinical Hypertension found that only 5% of emergency physicians would administer an antihypertensive for a blood pressure in the 180s.
“It was only when the systolic got to about 220 that you had 50% willing to administer,” Dr. Breu pointed out. That’s in part due to 2013 guidelines issued by the American College of Emergency Physicians that “are very clear,” he said. “They say, ‘Don’t treat.’ “