An effort to reduce antibiotic overuse in patients being discharged at 10 US hospitals did not achieve its primary goal, according to the results of a randomized trial published late last week in JAMA Network Open.
The trial, which was led by researchers at the University of Iowa Carver College of Medicine, found that a discharge-focused prospective audit-and-feedback (PAF) intervention on antibiotic consumption did not reduce antibiotic use at discharge when compared with the pre-intervention baseline. But optimal antibiotic prescribing for patients with common and uncomplicated infections was more common during the intervention.
The authors of the study say the findings suggest other strategies may be needed to reduce antibiotic use at discharge, which has long been seen as an important target for antibiotic stewardship. Research suggests that antibiotics prescribed when patients are discharged home from the hospital account for up at least 40% of all antibiotic days received by hospitalized adults in the United States, and up to 70% of those prescriptions are "suboptimal"— they either are unnecessary, taken too long, or use an overly broad-spectrum antibiotic.
Aiming to identify problematic postdischarge prescriptions
The trial, conducted from December 5, 2022, to November 17, 2023, involved 10 US acute care hospitals with antibiotic stewardship teams and supporting staff. Each stewardship team identified patients with common bacterial infections within their hospital—such as chronic obstructive pulmonary disease, skin and other soft-tissue infections, pneumonia, and urinary tract infections—as a focus for the intervention.
The three components of the intervention included developing or updating institutional guidelines for oral antibiotic step-down therapy, disseminating those guidelines to frontline prescribers, and performing PAF, which involved a real-time review of antibiotic prescriptions for patients who were expected to be discharged within the next 48 hours. Reviews were led by pharmacists or physicians designated as "stewardship champions," who provided feedback on the prescriptions they felt needed to be modified.
"Postdischarge antibiotics are frequently unnecessary or suboptimal," the study authors wrote. "Several nonrandomized trials have shown that auditing and providing feedback at discharge can improve antibiotic-prescribing at this transition of care."
To evaluate the effectiveness of the intervention, the researchers compared the frequency of postdischarge antibiotic prescribing during the intervention with the postdischarge prescribing rate established during a 24-week control period. Per the stepped-wedge cluster-randomized design of the trial, one hospital crossed into the intervention arm every two weeks after the baseline period.
Postdischarge antibiotics are frequently unnecessary or suboptimal.
Secondary outcomes included inpatient antibiotic use, length of hospital stay, and readmission. The researchers also reviewed the health records of 434 patients to assess optimal antibiotic prescribing—defined as a combination of appropriate antibiotic selection and duration—at discharge for patients who met certain criteria.
A total of 21,842 patient admissions (median patient age, 66 years; 61.3% male) were reviewed for the study, including 14,288 during the baseline period and 7,554 during the intervention period. The mean number of patients audited by hospital stewardship teams per week was 19.9, with roughly one-quarter of the audits resulting in feedback to prescribers.
Analysis of the primary outcome showed 21.8% of patients were prescribed postdischarge antibiotics during the intervention period, compared with 21.9% during the baseline period (odds ratio [OR], 0.94). The mean postdischarge antibiotic duration the intervention period was 7.6 days, compared with 7.1 at baseline. No statistical differences between the two periods were found for inpatient antibiotic duration, length of hospital stay, or readmission.
But the review of the electronic health records of patients with select infections found that optimal antibiotic prescribing at discharge increased from 46.2% at baseline to 58.5% during the intervention. An adjusted analysis found that those patients had significantly higher odds of receiving optimal antibiotics during the intervention period relative to the baseline (OR, 1.61, or 61% higher odds).
In a postintervention survey, 89% of respondents said the intervention should continue. But nine of the hospitals dropped the intervention once the trial ended.
Communication challenges
The authors suggest one reason the intervention wasn't successful at reducing postdischarge antibiotic use was that all the hospitals in the trial already had active antimicrobial stewardship (AS) discharge plans. But there could be other explanations.
"Second, many discharging patients may not have been touched by the intervention, either because their infection was too complicated or because an opportunity to optimize therapy was not addressed prior to discharge," they wrote. "Finally, frontline prescribers may have disagreed with the AS team’s input or simply forgotten to apply it when entering their discharge antibiotic orders."
In an accompanying commentary, experts from the University of Utah Eccles School of Medicine also suggest that the way PAF recommendations were relayed to prescribers during the intervention period could have played a role. They note that six of the 10 hospitals primarily used electronic messaging, rather than in-person or telephone feedback.
"Electronic communication does not allow for the subtleties of interpersonal interactions to be expressed, which may shape the acceptability of a recommendation to modify a prescription," wrote Julia Szymczak, PhD, Valerie Vaughn, MD, and Adam Hersh, MD, PhD. "Feasibility concerns with real-time in-person or telephone feedback may make it a less appealing option, but more research is needed to define optimal PAF implementation."
They added that more information is needed on the way that local adaptations to the intervention may have impacted its success.
