A large U.S. hospital study found that clinicians may have prescribed fewer broad-spectrum antibiotics when guided by electronic prompts in patients’ medical records.
These prompts appeared during antibiotic ordering among patients hospitalized with abdominal infections who were at low risk for drug-resistant bacteria.
The study involved over 198,000 adult patients across 92 hospitals. Investigators compared outcomes between hospitals using a computerized provider order entry (CPOE) system and those following standard antibiotic stewardship practices. The 12-month trial focused on non–critically ill adult patients who received antibiotics within the first 3 days of admission.
In the CPOE group, clinicians received real-time alerts if a patient’s estimated risk of multidrug-resistant organism (MDRO) infection was below 10%. The system then recommended switching to a standard-spectrum antibiotic, though clinicians could override the suggestion.
Use of broad-spectrum antibiotics dropped by 35% in the CPOE group—from 519 to 350 days of therapy per 1,000 empiric days. By contrast, the routine care group saw only a small decrease, from 519 to 500 days.
The prompts used more than 60 variables from patient records, including prior hospitalizations, infections, comorbidities, and local MDRO prevalence, to estimate individual risk. They aimed to support more targeted antibiotic decisions.
Reduced antibiotic use didn't worsen outcomes. The average hospital stay was 5.5 days in both groups. Intensive care unit transfer rates were also similar: 3.4% in the CPOE group and 3.6% in the routine group.
One key finding was that most patients had a very low risk of MDRO infection.
“The CPOE algorithm classified more than 98% of patients with abdominal infection in both groups as low risk; of these, less than 2% subsequently had an MDRO-positive culture,” said lead study author Shruti K. Gohil, MD, MPH, of the Division of Infectious Diseases at the University of California, Irvine School of Medicine, and colleagues.
Clinicians not only responded to the prompts, but also increasingly started patients on standard-spectrum antibiotics without prompting, reducing the need for alerts. This behavior suggested growing confidence in treating low-risk patients without broad-spectrum coverage.
Overuse of broad-spectrum antibiotics is a major driver of resistance. It can also increase the risk of complications, such as Clostridioides difficile infections, by disrupting the gut microbiome. More precise targeting of therapy may help minimize these risks.
The findings supported the value of real-time, personalized guidance in improving antibiotic prescribing, particularly for abdominal infections, where diagnosis is often delayed.
The trial was conducted across a national hospital system and supported by federal funding. It used an existing electronic health record platform, making the approach potentially scalable to other settings.
Full disclosures can be found in the published study.
Source: JAMA Surgery