A large observational study provides evidence that antibiotics provide no benefit for patients hospitalized with nonsevere COVID-19.
The study, published this week in JAMA Network Open, examined data on more than 520,000 US patients hospitalized for nonsevere COVID cases over a nearly 4-year period and found no clinically significant difference in outcomes between those who received antibiotics on day 1 of admission and those who didn't. In fact, patients who received antibiotics had slightly higher odds of poor clinical outcomes.
The authors of the study say the findings highlight the need for antibiotic stewardship strategies in patients admitted with nonsevere COVID-19.
High antibiotic use in COVID patients
Although COVID-19 is caused by a virus, at the beginning of the pandemic, more than 80% of hospitalized US COVID patients received antibiotics on admission, primarily because of limited treatment options and concerns about bacterial coinfections. There was also an early belief that azithromycin, in combination with hydroxychloroquine, might reduce COVID severity.
But retrospective data have since then shown that only 5% of COVID-19 patients had bacterial coinfections, and several randomized clinical trials (RCTs) have shown that azithromycin provides no benefit for COVID-19 patients. Once the first wave of COVID infections had passed and clinicians began using treatments like the antiviral drug remdesivir, monoclonal antibodies, and systemic steroids, antibiotic use in hospitalized COVID patients began to decline.
Even by the end of 2023, however, approximately 35% of US COVID patients were still receiving antibiotics on admission. The high rate of antibiotic prescribing in this population, combined with concerns about antibiotic resistance and potential patient harms from unnecessary antibiotic use, prompted researchers from the University of Wisconsin-Madison, the University of Massachusetts, and the University of Utah to assess clinical outcomes in hospitalized COVID patients who received antibiotics.
"When you look at the data and see that over 30% of COVID patients are still getting antibiotics, it's clear that they are still being widely utilized," lead study author Michael Pulia, MD, PhD, an emergency physician and associate professor at the University of Wisconsin-Madison School of Medicine and Public Health, told CIDRAP News. "So we have to get more data to show that it's either helpful or not."
When you look at the data and see that over 30% of COVID patients are still getting antibiotics, it's clear that they are still being widely utilized.
To do so, Pulia and his colleagues conducted a target trial emulation, which applies a randomized clinical trial framework to observational data in order to reduce the bias that typically occurs in observational studies. Using data from the Premier Healthcare Database, they identified adult, immunocompetent COVID-19 patients who were treated at US hospitals for COVID-19 from April 2020 through December 2023, excluding patients who had neutropenia, a non-pneumonia bacterial infection, or a chronic obstructive pulmonary disease (COPD) exacerbation.
"As the name implies, we were trying our best to replicate with observational data the conditions that are present at the time that somebody would be enrolled in a prospective randomized trial," Pulia explained. "And we had good, robust ways of excluding a lot of the people that wouldn't typically be eligible for a trial."
The researchers then assessed outcomes in patients who received a community-acquired pneumonia (CAP) antibiotic regimen or no antibiotics on day 1 of hospitalization, using propensity methods and adjusting for potential confounders.
The primary outcome was a composite measure of patient deterioration (vasopressor, high-flow oxygen, noninvasive ventilation, invasive mechanical ventilation, intermediate care, intensive care unit admission) and in-hospital mortality occurring on day 2 or later. The researchers considered an absolute standardized difference (ASD) of greater than 10% as demonstrating clinically meaningful differences between the groups.
No clinically meaningful difference in outcomes
A total of 520,405 patients (median age, 66 years; 51.2% male) treated at 1,053 US hospitals were included in the study. Of the patients, 160,482 (30.8%) were treated with a CAP antibiotic regimen on day 1 of admission.
Patients treated with antibiotics were more likely to be Hispanic or other race or ethnicity and more likely to be treated at hospitals with fewer than 400 beds, hospitals in the South or West, rural hospitals, and non-teaching hospitals. Overall, 95,055 patients (18.3%) deteriorated and 22,355 (4.3%) died during their hospitalization.
The analysis showed that the primary outcome was higher in the CAP group (20.8%) compared with the no-antibiotic group (18.4%), but the ASD (4.1%) did not meet the research team's predefined criteria for clinical significance. No clinically meaningful differences were found for secondary or safety outcomes.
But analysis of a matched cohort of 113,506 pairs using propensity score-weighted models found slightly higher odds of poor clinical outcomes associated with receipt of CAP antibiotics (propensity-matched odds ratio [OR], 1.03; 95% confidence interval [CI], 1.01 to 1.05; inverse probability treatment weighting OR, 1.03; 95% CI, 1.02 to 1.05; standardized mortality ratio weighting OR, 1.10; 95% CI, 1.08 to 1.12).
Pulia characterized that finding as a slight signal toward harm, but he said the larger point is that antibiotics did not appear to help the patients who received them.
"We basically found nothing to suggest in any of our results that antibiotics would be beneficial at all on the outcomes we looked at," he said. And if there's no benefit, he added, then the adverse impacts of antibiotic use on patients and public health need to be considered.
In a commentary, Sean Ong, MBBS, and Steven Tong, PhD, from the University of Melbourne, conclude, "Pulia and colleagues have demonstrated in their study an excellent example of how well-conducted observational research can fill existing research gaps where RCTs are lacking, and guide clinical decision-making." They add, however, that antibiotic therapy should be prescribed in COVID patients who were excluded from the analysis—such as those with COPD exacerbation—if there are clear indications for the drugs.
We basically found nothing to suggest in any of our results that antibiotics would be beneficial at all on the outcomes we looked at.
Pulia said that while prospective studies are needed, he hopes, at the very least, that the findings will make clinicians feel more comfortable in raising the threshold for initiating antibiotics in patients with nonsevere COVID, even if they are sick enough to be admitted to the hospital.
"Clearly the rate at which we're doing it today is too much," he said. "I think everybody's aware that it's a problem, it's just that it still happens a lot."
