Reducing Overuse of Antibiotics with Decision Support: The ROADS Study

PROJECT SUMMARY Lower respiratory tract infections (LRTI), including pneumonia, bronchiolitis, and infection-related exacerbations of asthma, are responsible for nearly one-third of all emergency department (ED) visits and 40% of all infection-related hospitalizations in children each year in the United States. While viruses are responsible for the vast majority of LRTIs, antibiotics are very frequently prescribed. In fact, more days of antibiotic use are attributable to LRTI than any other condition in children’s hospitals. Starting antibiotics unnecessarily, using less effective or overly broad-spectrum antibiotics, and treating for longer durations than necessary all contribute to inappropriate antibiotic use in pediatric LRTI in both ED and inpatient settings. Use of unnecessary antibiotics has profound consequences, including common and uncommon but severe adverse side effects, increased health care utilization and costs, and fueling the development of antimicrobial resistance, an ongoing global public health threat. To curtail inappropriate antibiotic use, Antimicrobial Stewardship Programs (ASPs) have become commonplace in United States hospitals and have proven to be effective. Unfortunately, the benefits of ASPs have largely failed to translate to the unique and fast-paced ED environment. This represents a key missed opportunity since many more children are cared for in the ED as compared to inpatient settings, and because ED care often critically influences subsequent hospital care. Our prior research suggests that high quality electronic health record-based clinical decision support (CDS) is a strategy that could bridge the ED antibiotic stewardship gap and reduce inappropriate prescribing in the ED and hospital. Our proposed research builds upon this foundational work and leverages an accomplished and diverse study team with expertise across the continuum of clinical care and in biomedical informatics, clinical and health services research, implementation science, and biostatistics, all housed within supportive and well- resourced research environments. Within three children’s hospitals, we will: define critical facilitators and barriers governing use and acceptance of antibiotic CDS at the point of care in the ED and inpatient settings (Aim 1); create two CDS interventions seamlessly integrated into clinician workflows in the ED and inpatient settings, including a) an antibiotic advisor (Abx CDS), and b) a real-time ASP tool (ASP CDS) (Aim 2); measure the effectiveness of Abx CDS and ASP CDS, alone and in combination, against usual care for increasing appropriate antibiotic use in pediatric LRTI in the ED and inpatient settings within the framework of a pragmatic randomized controlled trial (Aim 3); and evaluate Reach, Adoption, and Implementation domains within the trial among encounters receiving Abx CDS and ASP CDS (Aim 4). We hypothesize that we will create stewardship-focused CDS tools that meaningfully improve antibiotic use for children with LRTI and that are scalable in order to optimize care delivery and outcomes across a wide variety of settings.