Bronchiolitis is a viral respiratory illness affecting infants and toddlers and is the most common reason for hospitalization amongst children less than 2 years old. However, many children admitted with bronchiolitis receive unnecessary treatments that do not hasten recovery and may even cause harm. Despite evidence that the best treatment for bronchiolitis is supportive care (i.e. oxygen and hydration support), clinicians continue to overuse certain therapies, leading to longer hospital stays, higher costs, and increased stress for families.
The investigators will study the effects of EHR-alerts on clinical decision support (CDS) guideline use, and the effect of CDS guideline use on High Flow Nasal Cannula (HFNC) use and other clinical outcomes. EHR alerts are real-time alerts within the EHR which may be used for various purposes, but in this study are designed to remind providers that a detailed clinical decision support algorithm exists for the treatment of bronchiolitis. There are multiple BPA designs commonly used in EHRs: 1) interruptive alerts, which require clinicians to interact with the alert to continue the clinician's workflow; and 2) non-interruptive alerts, which appear as visual cues but do not require clinicians to alter workflows.
When patients meet criteria for the likely presence of bronchiolitis based on Electronic Medical Records (EMR) indicators of disease, the patient will be randomized into three groups in a 1:1:1 ratio. One group (control) will receive no EHR-alerts. In one group the providers will receive interruptive EHR-alerts reminding the provider that the CDS-guideline is available in the EHR and likely to be appropriate for the patient. In the last group, providers will be exposed to a non-interruptive alert (a highlighted textual banner presented to the provider when the providers are interacting with the patient's electronic medical record) reminding the provider that the CDS-guideline is available in the EHR and likely to be appropriate for the patient.
The randomized groups will be compared on an intention-to-treat basis to evaluate the impact of EHR-alert exposure on CDS-guideline usage as demonstrated by the number and nature of provider interactions with the CDS guideline tool. In addition, in a related non-randomized retrospective observational evaluation, we will compare the impact of CDS-guideline usage (the outcome of the randomized evaluation) on the provision of high-value care for bronchiolitis. Specifically, because the guideline emphasizes the use of high-value HFNC, and the avoidance of low-value HFNC treatments, this analysis with focus primarily on the use of HFNC in bronchiolitis. It will also evaluate whether CDS-guideline use impacts the rate of other low-value treatments for bronchiolitis, and outcomes such as ICU admission, length or stay and readmissions.