Using corticosteroids (or not)

Review the latest data on prescribing these drugs for pneumonia, sepsis, and COPD.


Whether systemic or inhaled, given or taken away, corticosteroids have the power to change a patient's course in the hospital and beyond, according to several speakers at Hospital Medicine 2019.

Photo by Stacey Butterfield
Photo by Stacey Butterfield

During an update on best practices in pneumonia management, ACP Member Joanna M. Bonsall, MD, PhD, urged hospitalists to put steroids to use when treating patients with severe pneumonia, pointing to the attractive number needed to treat to prevent mortality.

In a debate session, Daniel D. Dressler, MD, MSc, FACP, and Daniel J. Brotman, MD, FACP, reviewed the literature and ended up agreeing that steroids should be considered in severe pneumonia. They found less consensus on the utility of steroids in sepsis, although the two could agree on a small subset of patients who might benefit.

Finally, during an update on chronic obstructive pulmonary disease (COPD), Catherine Grossman, MD, said that removal of inhaled corticosteroids (ICS) may help prevent readmissions in certain patients admitted with COPD exacerbations. “There are those who need them,” she said. “I'm just asking you to apply it judiciously.”

Yes: Severe pneumonia

In recent years, many studies have assessed the effect of steroids in treating patients with community-acquired pneumonia.

One of the first large studies, published in October 2015 by Annals of Internal Medicine, was a meta-analysis of 13 trials comprising more than 2,000 patients. Patients had community-acquired pneumonia, and those with immunosuppression or side effects related to steroid use were excluded. “They found that doses of 20 to 60 mg of prednisone significantly lowered mortality in patients with severe pneumonia and also decreased length of stay and time to stability across the board,” said Dr. Bonsall, an associate professor of medicine and the director of hospital medicine at Grady Memorial Hospital in Atlanta.

Since then, additional meta-analyses have been published, and most have used stricter inclusion criteria than the Annals study, she noted. Of note, a Cochrane review, published in March 2017, again found a mortality benefit of steroids, with a number needed to treat of 18 to prevent mortality in patients with severe pneumonia. “They also found decreased time to stability and decreased length of stay in all comers,” Dr. Bonsall said. “I hope that that number needed to treat makes you all more comfortable with adding steroids, especially in your patients with severe pneumonia.”

Meanwhile, another meta-analysis published in the February 2018 Clinical Infectious Diseases “threw a wrench” in this theme, she noted. “What they found was that there was no impact on mortality, and for the first time, they found that there was increased readmissions, so that kind of made the headlines and gave everybody pause,” said Dr. Bonsall. However, the review only included six studies, and the increased readmission risk was confined to patients with low-severity pneumonia, she said.

There is little consensus on appropriate dose, so future trials will assess dosing regimens and durations, Dr. Bonsall said. “For my own personal practice, I typically pick somewhere from 40 to 60 mg and continue it for five days, but again, there's not a lot of evidence to tell us which way to go,” she said.

However, be sure to remember the exclusion criteria: patients with any pneumonia that is not community-acquired, immunocompromised patients, those with contraindications to steroids, and those with uncontrolled diabetes, Dr. Bonsall said, adding that this last criterion was not definitively established by research. “I think there's a little bit of wiggle room. . . . There was a smaller study that came out that looked at diabetic patients with an HbA1c as high as 11.6% and found that they also had the same mortality benefit,” she said.

During the steroids debate, Dr. Brotman (playing the con side) noted that hospitalists' typical pneumonia patients do not have severe disease. “So you should be asking yourself if your patients would have been randomized here,” he said. He did agree that the data provide adequate evidence for using steroids in patients with severe pneumonia.

Sometimes: Septic shock

Another subset of patients who may benefit from steroids are those with septic shock, said Dr. Dressler, professor of medicine and associate program director of the internal medicine residency program at Emory University School of Medicine in Atlanta. However, some physicians use steroids in sepsis while some don't. “The decision making is all over the place,” he said.

While the topic has been discussed for decades, it has become relevant again because of new, larger studies that were published last year, said Dr. Dressler.

The ADRENAL and APROCCHSS trials, both published in March 2018 by the New England Journal of Medicine, effectively doubled the number of septic shock patients in which steroids have been studied, he noted. While the former didn't find a difference in mortality with steroids, the latter found a significant reduction in 90-day mortality. A key difference was the mortality in the placebo groups: about 29% in the former and 49% in the latter, Dr. Dressler noted.

The trials led to a couple of meta-analyses as well as a clinical practice guideline from The BMJ. “That was a rapid guideline that was published after these meta-analyses . . . because they said this potentially could change practice,” Dr. Dressler noted.

The summary of the data is essentially that there may be a small benefit, he said, “But that benefit still is with mortality, so probably reasonably relevant.” Based on the data, he said his recommendation is to consider giving steroids in patients with refractory septic shock who are on vasopressors for more than six hours.

However, keep in mind that there is evidence of substantial publication bias in this literature, noted Dr. Brotman, also playing the con side for sepsis. The most definitive large study to date did not show a mortality benefit, and a portion of the meta-analytic signal came from small, poorer-quality studies that tended to favor steroids, he said.

“Steroids may help septic patients get more reassuring vital signs more quickly and maybe get out of out of ICU sooner, but it's not because you're perfusing their tissues better and making them better,” argued Dr. Brotman, professor of medicine and director of the hospitalist program at Johns Hopkins Hospital in Baltimore. “It's because you are improving their vital signs, and they may end up intubated again afterwards.” Ultimately, however, he said that considering steroids in the subset of patients with refractory septic shock and extended vasopressors is “probably a reasonable recommendation.”

Maybe: Maintenance inhalers for COPD

With the recent release of the Global Initiative for Chronic Obstructive Lung Disease (GOLD) 2019 report, hospitalists have fresh guidance on medication changes to consider in patients who have had a COPD exacerbation. Most patients they encounter are classified as GOLD group C or group D, noted Dr. Grossman, associate professor of medicine in the division of pulmonary and critical care medicine at Virginia Commonwealth University in Richmond.

In many of these patients, discontinuing an ICS might be a viable option. “It may be that the inhaled steroid is actually making you have more exacerbations by having more bacterial overgrowth, more bronchitis, more nonfatal pneumonias, and that may be contributing to your problems,” said Dr. Grossman.

To determine whether an ICS is needed, look for high blood eosinophil counts or try to find an overlap history of asthma and COPD, she said. “If this history doesn't exist, and that's likely not the majority of these patients, we can probably take them off, and it's been shown to be safe,” said Dr. Grossman, citing the WISDOM trial, published in October 2014 by the New England Journal of Medicine.

The GOLD guideline noted that multiple recent studies have shown that eosinophil counts predict the effect of ICS (added to regular maintenance bronchodilator treatment) in preventing future exacerbations. Minimal effects are observed at lower eosinophil counts, with incrementally increasing effects observed at higher eosinophil counts. In the GOLD recommendations, “Every place that moves you toward an inhaled steroid has some sort of eosinophilic qualifier,” Dr. Grossman noted.

One wrinkle in evaluating admission labs for eosinophilia is that emergency medical services (EMS) sometimes administer dexamethasone to patients prior to arriving in the ED. “Your admission labs may not show eosinophils if the patient already got a shot of dex from the EMS squad,” she said.

For patients who are taking a long-acting beta-agonist (LABA) or long-acting muscarinic antagonist (LAMA) and have a high eosinophil count (≥300 cells/µL, or ≥100 cells/µL plus a history of several exacerbations), a hospitalist could consider changing their controller medications to a LABA plus an ICS, Dr. Grossman said. Alternatively, for patients who don't fulfill these criteria, the hospitalist could try a LABA plus a LAMA, she said.

But what if a patient is on a LABA and ICS at the start? “I think a lot of primary care physicians still have patients on LABA plus ICS . . . for COPD,” Dr. Grossman said. “People come in, and my residents will say, ‘Well, that's just what they were on,’ and then I ask them to reflect, ‘Well, why were they on that? Put back on your thinking cap. Is that the right thing?’” In these patients without eosinophilia, one could consider a LABA and LAMA combination or, if eosinophilia is present, triple inhalers (LAMA plus LABA plus ICS).

The guidelines recommend starting GOLD group C and D patients with LAMA or LAMA/LABA combination inhalers, unless they are GOLD group D and have high peripheral eosinophil counts. In the latter case, the guidelines allow for starting an ICS/LABA combination inhaler.

If patients are already on triple therapy, a hospitalist could add roflumilast or a macrolide, the GOLD guidelines said. Alternatively, for patients with COPD on triple therapy with an ICS, there is always a treatment step that says, “Take out the ICS and see what happens,” Dr. Grossman said. “It may be causing you more problems than you think.”