No one knew what to expect with a new virus. So earlier this year, when the first U.S. patients were hospitalized in medical units and ICUs with COVID-19, their physicians utilized every possible tool in the armamentarium in an attempt to improve outcomes. One of those tools was antibiotic therapy.
Although everyone knows antibiotics don't work on viruses like SARS-CoV-2, they were frequently used due to concern about bacterial co-infection, said Valerie M. Vaughn, MD, MSc, FACP. “There was a lot of thought among many of us that, like with H1N1 and influenza, maybe the patients who were the most ill were those who came in with secondary infections,” she said.
As a hospitalist at the University of Michigan in Ann Arbor and the Veterans Affairs Ann Arbor Healthcare System, Dr. Vaughn led a study of antibiotic use in COVID-19 based on what she was seeing on the front lines. “I personally was prescribing a lot of antibiotics. So were my colleagues,” she said. “We didn't really know at the time how often co-infections occurred.”
Adding to the uncertainty, hospitals across the country were experiencing diagnostic testing delays. “Especially in early March, there were times when it was taking up to a week to get [SARS-CoV-2] test results back,” said Dr. Vaughn, who is now director of hospital medicine research and an assistant professor of medicine at the University of Utah in Salt Lake City. “So during that time, if you don't know for sure that the patient has COVID, you treat them like they have a bacterial infection, bacterial pneumonia.”
Between March 13 and June 18, more than half of 1,705 patients hospitalized with COVID-19 in 38 Michigan hospitals were prescribed early empiric antibacterial therapy, her study found. Overall, only 3.5% of the cohort had confirmed community-onset bacterial infection, according to results published online on Aug. 21 by Clinical Infectious Diseases. Testing delays did appear to be a factor—about 54% of patients had antibacterial therapy stopped within a day of a positive COVID-19 test result, and early empiric antibiotic use decreased as the turnaround time for test results decreased.
“Hospitalists and other physicians [should be] aware that co-infections are relatively rare when patients first come into the hospital [with suspected COVID-19]. . . . There aren't a lot of patients who actually need antibiotics up front,” Dr. Vaughn noted.
Fortunately, hospitalists have gained a greater sense of the novel coronavirus over time. In turn, unnecessary antibiotic prescribing in patients with COVID-19 may be improving, experts said. They offered advice for hospitalists on optimal antibiotic use in this population.
Problems with antibiotic overuse
Research from different parts of the world has found low rates of bacterial co-infection in patients with COVID-19.
In a study of 836 inpatients with confirmed SARS-CoV-2 infection at two London hospitals, 27 (3.2%) had confirmed bacterial isolates identified in the five days after admission. Throughout admission, this number rose to 51 (6.1%) patients, according to results published online on June 26 by Clinical Microbiology and Infection.
Another study of 4,267 COVID-19 patients hospitalized at Montefiore Health System in the Bronx, New York City, during the surge between March 1 and April 18 found that only 152 (3.6%) had bacterial or fungal co-infections. But out of 5,753 COVID-19 patients admitted between March 1 and May 31, 4,130 (71%) received at least one antibiotic dose, according to results published online on July 24 by Infection Control & Hospital Epidemiology.
The key takeaway is that, more often than not, giving antibiotics to a COVID-19 patient on admission is not necessary, said lead author of that study and infectious diseases subspecialist Priya Nori, MD, medical director of the antibiotic stewardship program for the department of medicine at Montefiore. “Now we have enough data to show that giving antibiotics up front and even continuously thereafter on an empiric basis does not seem to help, because most patients do not have concurrent bacterial or fungal infections at the time of their presentation with COVID,” she said.
Prescribing antibiotics when they are not necessary can have adverse consequences for both the patient and the population as a whole, said Dr. Nori, who is also an associate professor of medicine and of orthopaedic surgery at the Albert Einstein College of Medicine in the Bronx. Certain antibiotics can accelerate a patient's kidney failure, for instance, or lead to liver inflammation or drug rashes, she noted.
Another consequence of antibiotic overuse is Clostridioides difficile infection. Dr. Vaughn recalled caring for two patients with COVID-19 on a moderate-care floor who developed C. difficile infection during hospitalization. “The first fortunately recovered, but the second actually ended up having very, very severe C. difficile and then passing away from it,” she said.
Adding to the list of problems with antibiotic overuse, Dr. Nori noted that new drug development is failing to keep up with the rate of multidrug resistance. “Although we've made a lot of strides, there are still certain types of bacteria or fungi for which there are very limited antibiotic options, maybe one or two sometimes,” she said.
On Oct. 9, the U.S. government released its National Action Plan for Combating Antibiotic-Resistant Bacteria for the next five years. The report described how increased antibiotic use can complicate the response to and recovery from public health emergencies like the COVID-19 pandemic.
“While the implications of antibiotic resistance are not yet clear for the ongoing response . . . increased use of antibiotics and other antimicrobial medicines—both appropriate and inappropriate—to address primary or secondary infections has the potential to further accelerate the emergence of antibiotic resistance,” the report said.
Dr. Vaughn added that giving patients antibiotics early on could even increase their risk of having resistant organisms later. “Using them early makes them less effective later when you need them,” she said.
Dr. Nori confirmed that antibiotic resistance has posed problems in the treatment of the patients who do have a bacterial co-infection. “This is a very real phenomenon that we even saw playing out in some of our COVID patients because of the heavy exposure to antibiotics or the prolonged stay, whether it was in a long-term care facility or other health care facility,” she said.
The few patients with COVID-19 who go on to develop secondary infections during their hospital course are generally the ones who are most sick and most vulnerable (e.g., critically ill patients, those with septic shock), she noted. “You don't want to miss the secondary event that could be a bacterial or fungal infection, but there are probably certain things that can help predict whether that's happening,” such as higher fevers, oxygen requirements, or white blood cell counts, she said.
Stewardship tips and tricks
Every hospital is different. That's why the percentage of patients with COVID-19 who were prescribed empiric antibiotics varied from 27% to 84% in the Michigan study.
“Variation is very common in antibiotic prescribing all the time and, I think, related to stewardship resources and also the local culture,” said Dr. Vaughn. “There are places where everyone's used to thinking twice about their antibiotic use, and those places are less likely, I think, to have antibiotic overuse during a time of a pandemic.”
Hospitals that performed well in a prior initiative to reduce antibiotic treatment duration to five days for most patients with pneumonia also performed well with regard to empiric antibiotic use for COVID-19, she added, citing data not presented in the study. “I think it shows the need for continued stewardship and how stewardship in one area might affect the way we think about antibiotics in other areas,” said Dr. Vaughn.
Since 2017, every hospital accredited by The Joint Commission is required to have an antibiotic stewardship program, which can be an important resource during a pandemic, a recent article noted. Prospective audit with intervention and feedback and antibiotic “timeouts” are two strategies that can help optimize antibiotic use, according to the article, published online on July 3 by Current Infectious Disease Reports.
At least at some sites, antibiotic overuse in COVID-19 patients seems to be improving. At VCU Health in Richmond, Va., an early analysis found that use of community-acquired pneumonia-focused antibiotics increased in the first month of the pandemic on a step-down medicine unit and a medical ICU, said Michael Stevens, MD, MPH, FACP, director of the antimicrobial stewardship program and associate chair of the division of infectious diseases.
“In that second month, though, we had a reversion of our antibiotic consumption to what are baseline levels, so pre-COVID levels,” he said, attributing that success to a well-resourced antimicrobial stewardship program and a universal COVID-19 testing policy for all admitted patients that began in late April. “People felt more comfortable, once they got comfortable seeing these patients, waiting to start antibiotics, especially if ID was involved and saying, ‘Hold off on the antibiotics.’”
Stewardship experts can also support hospitalists by developing guidelines for their hospitals, said Dr. Stevens, who coauthored a letter to the editor with Dr. Nori about ways stewardship programs can help during a pandemic, published online on March 13 by Infection Control and Hospital Epidemiology. “One of the things that we came up with was treatment guidelines. So locally, I said, ‘All right, we can't write this and not do it,’” he said.
After working with different programs across the country, and also looking at guidelines from Italy and China, experts at VCU Health created local treatment protocols, including guidance on empiric antibiotics for possible bacterial co-infection, Dr. Stevens said. “This thing started out as six pages long. We released them on March 15. Since then, we've updated them over 65 times, and it's a 24-page-long document where we critically assess all the new literature that comes out,” he said. Disseminated through the institution's intranet and mobile app, the COVID-19 treatment guidelines (which mainly consist of tables and flowcharts) are referenced more than 500 times a month on average, Dr. Stevens added.
At Montefiore, Dr. Nori's program also created a local treatment guideline and in early May put together a set of empiric antibiotic guidelines for COVID-19 patients to help clinicians diagnose and manage secondary infections. “We did this because we felt the need. We didn't see this coming out in any literature, there were no guidelines from the NIH or our professional societies. No one was talking about the antibiotic part,” she said.
Dr. Vaughn added that at a time when stewardship is more important than ever, stewardship programs are also more taxed and under pressure than ever. “Many have had cuts in resources and now have more responsibilities, such as remdesivir allocation or guideline creation,” she said. “It's a setup for problems, and we really are going to need to work hard to make sure that COVID-19 doesn't lead to more antibiotic resistance.”
In addition, antibiotic overuse in patients with COVID-19 hasn't improved everywhere. Across Michigan hospitals, updated data showed that the use of empiric antibacterial therapy in COVID-19 patients has plateaued, Dr. Vaughn reported in correspondence published online on Oct. 10 by Clinical Infectious Diseases. The proportion of patients receiving early empiric antibiotics declined from 67.8% in March to 47.5% in April before hovering around 50% through August.
“COVID numbers now are upticking, and antibiotic use is still an issue,” she said in late October. “Thus, it is imperative antibiotic stewardship programs continue to help identify patients for whom early empiric antibiotic therapy is not indicated and focus instead on directing treatment toward those with signs of secondary, hospital-acquired infections.”