Interruptive alert triggered by nonoptimal antibiotics for diagnosis of acute bacterial sinusitis
Henry Ford Health, Detroit, MI*
Dowload PDF
Last Updated
September 29, 2026
Populations impacted
Ambulatory adults
Description
Identifying target disease states that are ideal for stewardship intervention can be challenging. The Centers for Disease Control and Prevention Core Elements for Antibiotic Stewardship suggest identifying high-priority conditions where clinicians deviate from best-practices in antibiotic prescribing as targets for antimicrobial stewardship program intervention. The Henry Ford Health ambulatory ASP’s internal prescribing data suggested that acute sinusitis is a top five indication for an antibiotic prescription and that most are suboptimal regarding drug selection and duration of therapy. In response, the HFH ambulatory ASP developed an interruptive electronic health record hard-stop alert designed to improve drug selection and duration of therapy for acute bacterial sinusitis diagnosis for patients seen in ambulatory settings (i.e., clinics, urgent care, virtual visits). The interruptive acute bacterial sinusitis alert received support from our health system and ambulatory ASP committees and was implemented April 3, 2025.
The HFH ambulatory ASP leveraged information technology specialists to develop a forward-facing alert that follows best practices in effective design. The alert was developed along with an otolaryngologist champion, and consideration was given to ambulatory provider workflows. The alert activates when nonoptimal antibiotics or durations are prescribed for bacterial sinusitis encounters and provides a single-click prompt to utilize an antibiotic order sentence embedded with optimal antibiotic selection (i.e., amoxicillin-clavulanate or doxycycline) and defaulted with a best-practice 5-day duration. Additionally, the ambulatory ASP educated primary care providers about the alert through an electronic handout and during leadership meetings.
We anticipated the alert would significantly improve optimal antibiotic prescribing in acute sinusitis. While most acute sinusitis is not bacterial in nature or reflects other etiologies (i.e., headache/migraine, allergy, reflux or post-nasal drip), the HFH ambulatory ASP sought to mitigate suboptimal antibiotic prescriptions given there are roughly 30,000 patients who receive an antibiotic prescription for acute sinusitis at our health system each year. Our program has several other initiatives in place to promote optimal sinusitis diagnosis, which represents the underlying problem of antibiotic overuse.
Impact and data
The alert was formally evaluated by clinical pharmacy, infectious diseases, and primary care team members and led by a clinical pharmacy fellow. An institutional review board-approved, single pre-, post-test quasi-experiment was performed to evaluate optimal sinusitis prescribing, defined as amoxicillin/clavulanate or doxycycline (β-lactam allergy) for a 5-day duration, as recommended by institutional guidelines for sinusitis treatment.
A total sample of 200 patients were evaluated: 100 pre-intervention, 100 post-intervention. The primary outcome, the prescribing of optimal antibiotic therapy, occurred in 7% of patients in the pre-intervention group and in 30% of patients in the postintervention group (unadjOR, 5.69; 95% CI, 2.36-13.72; P<0.001). The median (IQR) duration of therapy in the preintervention group was 7 (5-10) days and 7 (5-7) days in the postintervention group (P=0.004). The median (IQR) time to acknowledge the interruptive alert was 7 (4-15) seconds. The alert was bypassed 58 times in the postintervention period; 47 instances resulted in nonoptimal antibiotic prescribing that primarily represented suboptimal duration. There was also an observed reduction in interruptive alert activations over the postintervention time frame.
An additional step the HFH ambulatory ASP is taking is to more formally evaluate patient characteristics or other confounding variables that could be associated with optimal sinusitis antibiotic prescribing. Additionally, a preliminary review of the EHR hard-stop alert performance suggests >25% acceptance as of January 2026.
Smith MR, Lee JJ, Holubar M, et al. Clinical decision support in the electronic health record: a primer for antimicrobial stewards and infection preventionists: work smarter so end users don’t work harder. Antimicrobial Stewardship & Healthcare Epidemiology. 2024;4(1):e204. doi:10.1017/ash.2024.448
Henry Ford Health Interruptive Alert Sinusitis Poster (PNG)
Henry Ford Health Sinusitis Alert Screenshot (PNG)
Implementation tips
Ambulatory ASPs should consider a local evaluation of sinusitis antibiotic prescribing to assess if a similar interruptive EHR hard-stop alert could be applicable. Our alert was developed with complementary expertise and in conjunction with designated analyst support. It was important to obtain perspective from the ambulatory providers who write antibiotics for sinusitis when designing the alert; other programs should consider contact tracing to maximize alert effectiveness within provider workflows.
The HFH ambulatory ASP is considering an audit and feedback approach for providers who do not accept the alert based on sinusitis prescribing volume. We have found that providers often use previously saved “preference lists” that can be outdated with antibiotic selection and duration. These are lists the provider can create themselves or share with other colleagues or trainees. Many suboptimal antibiotic prescriptions are generated from these preference lists. Ideally, our ambulatory ASP would be included in the oversight of preference lists to promote optimal antibiotic use.