Broad Implementation of Outpatient Stewardship (BIOS)

American Academy of Pediatrics, University of Pennsylvania, The MetroHealth System, Penn State Health, Nationwide Children's Hospital, Pediatric Associates of Florida, Patient-Centered Outcomes Research Institute

Reducing inappropriate antibiotic use is a key strategy to mitigate antibiotic resistance and adverse health effects associated with antibiotic exposure. The Broad Implementation of Outpatient Stewardship (BIOS) project focuses on broadly implementing an evidence-based intervention to improve antibiotic prescribing for acute respiratory tract infections in pediatric outpatient settings. Primary aims include: (1) examining the acceptability, feasibility and utility of a focused implementation strategy on improving intervention adoption and impact and (2) measuring the effectiveness of the intervention to reduce unnecessary broad-spectrum antibiotic prescription.

Antibiotics are commonly prescribed for acute respiratory tract infections (ARTIs) in pediatric outpatient settings, but up to half of antibiotic use is inappropriate. Prior work demonstrated broad-spectrum antibiotics did not improve patient health outcomes compared to narrow-spectrum antibiotics, but did increase harmful side effects. Overuse of broad-spectrum antibiotics can exacerbate antibiotic resistance and drug-related adverse events. Certain interventions have been effective in improving antibiotic prescribing, but none have been implemented widely. The BIOS project focuses on broadly implementing an evidence-based intervention to improve how clinicians in outpatient settings prescribe antibiotics for ARTIs in children 6 months to 12 years old. The intervention consists of educational modules and prescribing audit and feedback reports delivered to clinicians in a variety of outpatient settings across 5 health systems. Primary aims include: (1) examining the acceptability, feasibility and utility of a focused implementation strategy on improving intervention adoption and impact and (2) measuring the effectiveness of the intervention to reduce unnecessary broad-spectrum antibiotic prescription. Clinicians will be randomized to one of two arms: an early intervention arm or a delayed intervention (control) arm. The study will use a 4-period design, where the periods are as follows: Period 0: Baseline period that occurs prior to randomization Period 1: clinicians in the early intervention arm receive the intervention Period 2: All clinicians (both arms) receive the intervention Period 3: Maintenance period, external support from the study team is removed

The delayed intervention (control) arm will begin receiving the intervention in study period 2. They will receive no intervention during period 1.

The BIOS intervention consists of 3 components: (1) online educational modules covering topics related to appropriate antibiotic prescribing for ARTIs, (2) individualized audit and feedback reports displaying clinicians' rates of antibiotic prescription with peer comparison, (3) facilitation provided by the study team and by health system champions to support effective implementation of the intervention and to encourage clinicians to engage meaningfully in the intervention activities.

The proportion of visits for bacterial ARTIs during which a broad-spectrum antibiotic was prescribed. Assessed using electronic health record data.

from baseline (up to 36 months prior to randomization) through each of the study periods (up to 27 months of intervention time following intervention initiation)

The proportion of visits for viral ARTIs during which any antibiotic was prescribed. Assessed using electronic health record data.

from baseline (up to 36 months prior to randomization) through each of the study periods (up to 27 months of intervention time following intervention initiation)

Investigators will measure clinicians' completion rates of the educational modules and view rates of the feedback reports. Assessed through clinician attestation.

Investigators will use surveys and qualitative interviews to measure the extent to which the intervention was implemented as designed.

Surveys administered at the start of or just prior to period 1, during period 2, and during period 3. Interviews conducted during period 2.

The proportion of visits for bacterial ARTIs during which a broad-spectrum antibiotic was prescribed by practice setting (Emergency department, urgent care, pediatric primary care, family medicine/medicine-pediatrics). Assessed using electronic health record data.

from baseline (up to 36 months prior to randomization) through each of the study periods (up to 27 months of intervention time following intervention initiation)

The proportion of visits for bacterial ARTIs during which a broad-spectrum antibiotic was prescribed by ARTI type. Assessed using electronic health record data.

from baseline (up to 36 months prior to randomization) through each of the study periods (up to 27 months of intervention time following intervention initiation)

Rate of antibiotic prescribing for viral ARTIs (for which antibiotic therapy is not indicated) by practice setting

The proportion of visits for viral ARTIs during which any antibiotic was prescribed by practice setting (Emergency department, urgent care, pediatric primary care, family medicine/medicine-pediatrics). Assessed using electronic health record data.

from baseline (up to 36 months prior to randomization) through each of the study periods (up to 27 months of intervention time following intervention initiation)

The proportion of visits for all ARTIs (viral and bacterial) during which an antibiotic was prescribed

from baseline (up to 36 months prior to randomization) through each of the study periods (up to 27 months of intervention time following intervention initiation)

The proportion of visits for all ARTIs (viral and bacterial) during which an antibiotic was prescribed by practice setting (Emergency department, urgent care, pediatric primary care, family medicine/medicine-pediatrics). Assessed using electronic health record data.

from baseline (up to 36 months prior to randomization) through each of the study periods (up to 27 months of intervention time following intervention initiation)

Inclusion Criteria: Clinician inclusion criteria: Any non-trainee clinician who prescribes and provides care for children aged 6 months to 12 years with ARTIs and who is practicing at an included site. Clinicians will further be selected based on employment status at time of randomization and volume of prescribing over the past 12 months. Visit inclusion criteria: visits made by patients of all participating practices aged 6 months through 12 years of age will be considered part of the project population. Exclusion Criteria: Clinician exclusion criteria: clinicians who are trainees, who do not prescribe (or prescribe very infrequently) to the patient population of interest, or who do not provide care for (or provide care very infrequently for) children aged 6 months to 12 years with ARTIs will be excluded Visit exclusion criteria: visit by children with complex chronic conditions, as well as those treated with an antibiotic in the past 30 days, will have their data extracted but will be excluded from analyses