Published in the January 2017 issue of Today’s Hospitalist
IT’S ONCE AGAIN PRIME TIME for pneumonia, and hospitalists around the country are deciding how best to make that diagnosis and treat the condition. But for doctors who’ve gotten into a routine in their approach to pneumonia, it may be time to rethink some basic clinical strategies and options. That’s according to Scott Flanders, MD, a pneumonia guru who is hospital medicine director at the University of Michigan Health System in Ann Arbor.
In his pneumonia update at the management of the hospitalized patient conference held last fall by the University of California, San Francisco, Dr. Flanders started his talk by stressing the need to fine-tune pneumonia treatments, particularly antibiotics.
“UTI and pneumonia account for over half of all inpatient antibiotic use,” he pointed out. “In talking about pneumonia, the onus is on me to focus on rational antibiotic use.”
What constitutes rational antibiotic use for pneumonia these days? According to Dr. Flanders, the answer is to choose the right drugs (which might be none at all) for the right patient at the right time, and to keep the duration short. To back up the need to scale back on antibiotics, he pointed to a recent CDC study in the November 2016 issue of JAMA Internal Medicine that analyzed 550 U.S. hospitals over six years. Among inpatients, 55% received antibiotics and, increasingly, broad-spectrum therapies.
“In my hospital, 25% of E. coli is resistant to fluoroquinolones.”
“Broad-spectrum antibiotics save lives,” he said. “But multiple studies have shown that between 30% and 50% of antibiotics in the hospital are potentially inappropriate.” Another CDC study, which appeared in the July 30, 2015, issue of the New England Journal of Medicine (NEJM), looked at 2,000 pneumonia-diagnosed inpatients or controls in five large hospitals who were screened and extensively tested to determine the pneumonia etiology. Seven out of 10 were viral, only three were bacterial—and only 4% were found to have pneumococcus.
“When I reflect on how many patients we actually treat with antibiotics when they come in with pneumonia, it’s roughly 100%, plus or minus zero,” said Dr. Flanders. “In 90% of patients, we can’t find any organism or it’s a virus, but we treat 100% of patients for bacteria. That is is a big disconnect for me.”
CT, XR or both?
In terms of diagnosing pneumonia, Dr. Flanders posed a provocative question: Is the chest X-ray still the gold standard, now that CTs—which have been shown to aid in diagnosis—are readily available in most institutions?
It’s a tough question, in part because X-ray and CT findings often differ. For instance, a French study published in the October 15, 2015, issue of the American Journal of Respiratory and Critical Care Medicine that looked at suspected pneumonia patients in the ED had disconcertingly conflicting findings. Researchers found that almost one-third of the 120 patients with a negative X-ray “had a CT scan that demonstrated infiltrates consistent with pneumonia.” And in 190 patients with a CXR showing pneumonia, almost one-third had infiltrates on X-ray but not on CT.
Based on CT results, 61% of the emergency physicians to whom patients had presented indicated that they would have changed their treatment plan, with 15% saying they would have stopped antibiotics and 40% claiming they would have started them. The authors suggested that CT be routinely used, and Dr. Flanders noted that some institutions have “latched onto the study” to jump on the routine-CT bandwagon.
But he doesn’t agree. “It goes without saying that these authors did not look at clinical outcomes with this strategy, so we don’t know if it leads to better care,” he said. “It leads to different care, not necessarily better care.”
Further, “CT results differed in this study when there was already a clinical-chest X-ray disconnect,” he said. For example, an elderly patient with a negative X-ray who looks sick and has an abnormal lung exam and a high white-cell count might warrant treatment based on the hospitalist’s clinical judgment—without having to order a CT.
“My takeaway from this study is that maybe we should trust our clinical judgment more in cases where there are disconnects between how the patient looks and the X-ray, and focus on our clinical diagnosis,” he said.
Which antibiotics, when?
When patients do have pneumonia and need antibiotics, what’s the best regimen? Dr. Flanders focused first on the controversial subject of when or why macrolides are in order.
The current thinking, based in part on a meta-analysis published in the August 2012 Clinical Infectious Diseases, is that any regimen that includes a macrolide is associated with better outcomes. While the drugs don’t have “much effect on patients shown to have atypical organisms,” he noted, “macrolide-based treatment is associated with significantly reduced 30-day mortality.”
What’s the basis for that effect? The hypothesis, he explained, is that macrolides likely have immunomodulatory properties—and that they’ll emerge, like statins, as drugs that provide a benefit beyond their intended purpose. Experts haven’t yet figured out exactly why macrolides are so beneficial, and the current data come primarily from observational studies. Still, those data increasingly support the drugs’ efficacy in treating pneumonia.
“That’s why macrolides figure so prominently in guideline recommendations,” he noted. “They’re up top for our ward patients: a beta-lactam plus a macrolide.” Doxycycline is the leading alternative followed by a respiratory fluoroquinolone, but the beta-lactam-plus-macrolide regimen is also the leading recommendation for the ICU setting. (Another guideline-recommended approach is combining a fluoroquinolone with the beta-lactam.)
But while the beta-lactam plus macrolide sounds like the best way to go, Dr. Flanders has problems making a firm recommendation. He noted that there are conflicting data on macrolides, including randomized trials that found no benefit in those regimens over non-macrolide ones.
“We haven’t really had the trials that looked at the questions we’re all asking,” he said. “Do I add a macrolide to a beta-lactam, and do I need to do ceftriaxone plus something else, or can I just do ceftriaxone alone?”
Macrolides: not a slam dunk
Fortunately, two recent trials can help. One, a Swiss randomized trial that included 580 community-acquired pneumonia (CAP) patients, compared beta-lactams plus macrolides to beta-lactams alone.
The study found that patients receiving macrolides were more likely to be clinically stable at seven days than those who didn’t: 67% vs. 59%. (Results were published in the December 2014 JAMA Internal Medicine.)
The macrolide regimen was particularly beneficial in patients with atypical organisms and those who were relatively sicker when therapy began. Adding a macrolide, however, didn’t affect ICU admissions, length of stay, complications, recurrence or mortality.
A second, more complicated noninferiority trial was published in the April 2, 2015, issue of NEJM. Researchers randomized hospitals caring for about 2,000 non-ICU patients without legionella to rotating antibiotic regimens of beta-lactam monotherapy, beta-lactam plus macrolide or fluoroquinolones.
The trial concluded that beta-lactam monotherapy was noninferior. Interestingly, the fluoroquinolone arm had the lowest mortality (8.8%), while the dual-therapy macrolide regimen had the highest (11%). Length of stay was six days in all three arms.
“Some of you may be saying, ‘Let’s just give a macrolide to everybody and be done with it,’ ” said Dr. Flanders, who admitted feeling “a little bit of the same.” But on the flip side, “I think we have to be cautious.”
Macrolides’ big potential downside is excess cardiovascular mortality, presumably due to QT prolongations. Although the absolute risk is not very high—approximately one incidence in every 5,000 treatment courses—hospitalists should keep in mind that many hospitalized patients receive other QT-prolonging medications.
“We tend not to monitor those patients on a large population scale,” he pointed out. “If you’re making a recommendation for what everybody across the country should do, all of a sudden that CV mortality becomes relevant.”
He said he recommends a beta-lactam plus macrolide for sicker patients. But for those less sick, adding a macrolide might not be necessary.
Many hospitalists likely use fluoroquinolones instead, avoiding the whole macrolide debate. But, Dr. Flanders said, doctors should keep the downsides of that therapy in mind.
The big problem: Both pseudomonas and E. coli are increasingly resistant to fluoroquinolones, which have also been implicated in rising C. diff rates. In addition, adverse drug events with fluoroquinolones are frequent enough that the FDA recently warned against using them in uncomplicated infections.
“In my hospital,” said Dr. Flanders, “25% of E. coli is resistant to fluoroquinolones.”
Avoiding antibiotic overkill
Another major point that doctors need to rethink in treating pneumonia: Keep antibiotic duration short. Most guidelines still call for between 10 and 14 days of treatment and a five-day minimum, but that’s overkill, Dr. Flanders noted.
Studies increasingly show that shorter treatment is both effective and beneficial. One randomized study in the September 2016 JAMA Internal Medicine found that patients who received only five days of antibiotics and were afebrile for two days before treatment was stopped fared as well as those treated for longer periods.
Other antibiotic tips: If C. diff is an issue at your hospital, consider a beta-lactam plus doxycycline. Dr. Flanders cited an observational trial published in the September 2012 issue of Clinical Infectious Diseases that looked at the use of ceftriaxone plus doxycycline vs. regimens that did not include doxycycline. The research found a 27% lower rate of C. diff for every day of doxycycline.
He also recommended steering clear of drugs that put patients at high risk for C. diff. These include clindamycin, fluoroquinolones, and cephalosporins or carbapenems. Penicillins, macrolides and trimethoprim/sulfamethoxazoles are lower risk, and tetracyclines pose little risk.
And most patients with aspiration pneumonia do well with CAP treatment. Most pneumonia is due to aspiration—usually micro-aspiration—of oropharyngeal bacteria, Dr. Flanders added, and it rarely requires big-gun add-ons. If you need to specifically target anaerobes, consider ceftriaxone plus metronidazole. In a study of nursing home patients with anaerobes published in the March 2014 issue of the Journal of Internal and Emergency Medicine, 85% improved without dedicated anaerobic treatment.
Patient subsets who might need anaerobe-targeted coverage include individuals with severe periodontal disease, putrid sputum, necrotizing or cavitary lesions, large effusions, or alcoholism. In addition, patients who’ve had days or weeks of progressively worsening symptoms are more likely to have anaerobic infection.
Just say no to steroids?
For hospitalists wrestling with if or when to add a steroid, Dr. Flanders urged holding off in all but the most severely ill pneumonia patients. A meta-analysis of 13 randomized trials published in the October 2015 Annals of Internal Medicine found that adding a steroid offered little benefit in run-of-the-mill CAP patients. It did, however, reduce mortality (3%) and the need for mechanical ventilation (5%), although steroids also increased hyperglycemia risk by 3.5% or more.
“Right now, the answer to the question ‘Should we be using steroids?’ is ‘no,’ ” Dr. Flanders said. “The benefit is almost universally demonstrated only in the severe pneumonia group. And most of the mortality benefit is driven by one very small study that nobody has been able to replicate.”
He also pointed out that studies to date have included such a wide range of steroid doses and duration that it’s hard to tease out the right regimen.
“I have no idea who you should treat with steroids, how you should treat them, or how long and at what dose,” he said. On the plus side, he noted, several large trials are now under way to answer those questions.
Bonnie Darves is a freelance health care writer based in Seattle.
Procalcitonin: diagnostic darling, or downside?
ONE OF THE HOTTEST topics in pneumonia diagnosis right now is the potential role of procalcitonin, an inflammatory mediator that increases with microbial toxins, more so with bacterial infections than with viral ones.
Five years ago, few institutions were using the test, but now it’s widely available, said Scott Flanders, MD, a pneumonia expert who is hospital medicine director at the University of Michigan Health System in Ann Arbor. Because patients’ procalcitonin levels go down with clinical response, the test is gaining favor as a prognostic tool and a potential marker for when to safely discontinue antibiotics.
Should hospitalists use the test routinely? In Dr. Flanders’ view, the test can be helpful if properly used, but hospitalists shouldn’t rely on it too heavily to determine care. A November 2013 study in the Journal of Infectious Diseases found that some patients with bacterial pneumonia had almost undetectable procalcitonin levels, while levels were elevated in almost one-quarter of those with viral pneumonia etiologies.
“This is not a perfect test,” Dr. Flanders noted, “It should not be used as the sole variable to decide whether to give antibiotics or not.”
If hospitalists or their institutions do decide to use procalcitonin, they need to do serial testing, as recommended—but apparently, they’re not. According to Dr. Flanders, a study done in Michigan hospitals that’s awaiting publication found that 85% of patients tested for procalcitonin were tested only once.