Hypotheses and Specific Aims:
My prior work in croup demonstrated that more testing and treatment does not improve patient outcomes,1and that a bundled de-implementation strategy can significantly reduce over-testing and over-treatment in patients with bronchiolitis. In one year, at our hospital we reduced use of chest x-rays by 40%, and bronchodilators by 41% without negative clinical consequences for patients.2Moreover, there was sustained and continued improvement in years two and three, and a significant “spill over” effect in reducing antibiotic use, likely related to decreased chest x-ray use minimizing over-diagnosis of pneumonia. Widespread dissemination of this bundle would have significant clinical impact. However, a gap remains in our understanding of whether the influencing factors targeted by our bundle are generalizable across settings, and whether our bundle can be feasibly implemented in other settings. This knowledge gap limits our ability to spread the de-implementation strategies to other settings and diseases. To address this gap, the proposed project will employ a multiple level, theory-driven approach to 1) identify provider-, patient-, and setting-level factors influencing overutilization in bronchiolitis, 2) use these data to iteratively develop a targeted menu of strategies to de-implement tests/treatments (called Reduce Over-Utilized Tests and Treatments in Bronchiolitis or ROUTT-B), and 3) utilize a novel learning health system (LHS) called PEDSnet to conduct a feasibility study of these de-implementation strategies for reducing unnecessary tests/treatments in bronchiolitis patients. The overarching goal of this work is to advance the field of de-implementation science by identifying de-implementation processes and strategies that could be broadly adapted to varied contexts and diseases.
Aim 1: Identify factors influencing the overuse of chest x-rays, bronchodilators, and viral testing in bronchiolitis.At one high overuse and one low overuse PEDSnet institution, using PRISM(Practical, Robust Implementation and Sustainability Model)8as a guide, I will use a combination of qualitative measurement techniques such as semi-structured interviews and mini ethnography to define1) organizational-, 2) patient-,and 3) provider-level influencing factors either driving or passively facilitating the overuse of tests/treatments at each site, as well as the acceptability and feasibility of potential de-implementation strategies.
Aim 2: Develop ROUTT-B –a menu of pragmatic, feasible, and effective de-implementation strategies for over-utilized tests and treatments in bronchiolitis that includes guidance on how to choose strategies based on an assessment of local influencing factors. I will work with a group of implementation science and quality improvement experts to match potentially actionable provider-, patient-, and setting-level factors with evidence-based behavior change techniques and system level de-implementation strategies.
Aim 3: Conduct a pilot study to determine the feasibility, acceptability, and de-implementation effectiveness of ROUTT-B. Using PEDSnet, I will conduct a 1-year non-randomized controlled pilot trial at one high (intervention) and one low (comparison) overuse PEDSnet hospital to measure stakeholder perceptions of feasibility and acceptability of ROUTT-B and preliminary de-implementation effectiveness (rates of CXR, bronchodilators, and viral testing).