The rise of antibiotic stewardship efforts has seen success, but evidence is heterogeneous across different conditions. Intra-abdominal infection is one such area where there is a relative paucity of data for effective stewardship interventions. Given the diverse set of potential abdominal pathogens, the possibility of quickly developing critical illness, and the uncertainty of the underlying causes, we often reach for broad-spectrum empirical antibiotics with prolonged courses in patients with intra-abdominal infection.
The Study to Optimize Peritoneal Infection Therapy (STOP-IT) trial was a leap forward in antibiotic stewardship for intra-abdominal infection to limit the duration after source control was achieved.1 The current article by Gohil and colleagues2 takes the next step to address overuse of broad- or extended-spectrum antibiotics in this population. The authors conducted a pragmatic cluster randomized trial comparing computerized physician order entry (CPOE) clinical decision support vs routine antibiotic stewardship education. They developed a robust prediction model to determine from electronic health record information whether individual trial patients had high (>10%) risk of multidrug-resistant organism (MDRO) infection. For low-risk patients, the CPOE decision support prompt would fire, recommending standard-spectrum antibiotics; this ultimately showed a significant reduction in use of extended-spectrum antibiotics without increasing hospital or intensive care unit days.
This trial adds significantly to the literature and our understanding of effective interventions for antibiotic stewardship in this heterogeneous group of patients. The use of CPOE decision support systems has shown success, especially when configured to automatically provide individualized recommendations to clinicians, as the authors have done.3 However, there are several potential pitfalls to relying on CPOE decision support, such as workflow disruption, alert fatigue, and inappropriate or incorrect data incorporation that can derail the intended effects.4 For instance, many of the predictors of high MDRO risk were a history of such organisms or prior abdominal surgery, relying on data present in the electronic health record or clinician input. This information may not be available readily available in all patients early in their course when selecting empirical coverage, particularly those transferred from other institutions that do not share a medical record and thus may have an underestimated risk of MDRO abdominal infection.
Now that Gohil and colleagues have shown efficacy of their intervention, the next big question is how well it will scale to be implemented broadly. While they enrolled several hospitals, each were part of a single health care system with the same CPOE system. Work to understand how to translate and adapt this intervention across electronic health records and other health care systems with individual contexts will be important. Evaluating the sustainability of such an intervention over time is also critical to ensure we can capitalize on reducing the harms associated with unnecessary extended-spectrum antibiotic use in low-risk patients with intra-abdominal infection.
Published Online: April 10, 2025. doi:10.1001/jamasurg.2025.1107
Conflict of Interest Disclosures: None reported.
Leave a Reply