An Electronic Phenotype Series for Measuring Adherence to the Wake Up and Breathe Protocol in Mechanically Ventilated ICU Patients: Development and Validation Study (Preprint)

Le résumé fourni par la source

BACKGROUND The Wake Up and Breathe protocol integrates daily spontaneous awakening and breathing trials to facilitate earlier weaning from mechanical ventilation, improving outcomes for patients with acute respiratory failure. Measuring adherence to this multistep, interprofessional practice is essential but challenged by informatics complexity and risks of metric-specific gaming behavior. OBJECTIVE To define and validate an electronic phenotype series that enables retrospective, scalable, and reliable measurement of protocol adherence using real-world electronic health record (EHR) data. METHODS We derived structured, sequential phenotype definitions based on clinical guidelines and implementation resources, with a patient-day as the unit of analysis. The definitions were operationalized using heterogeneous data elements from a multi-hospital EHR warehouse. Phenotypes captured eligibility criteria, key process steps, and clinical outcomes. Manual chart review of 200 stratified patient-days was conducted to validate the phenotypes, with agreement measured using Cohen’s Kappa. RESULTS Seven phenotypes were defined, including two for upstream eligibility (@ReceivingIMV, @WeanEligible), four for protocol process steps (@SATEligible, @SATReceipt, @SBTEligible, @SBTReceipt), and one for an outcome measure (@SustainedExtubation). Agreement with chart review ranged from moderate (e.g., κ = 0.53 for @SATReceipt) to near-perfect (e.g., κ = 0.83 for @SATEligible), with lower agreement linked to documentation granularity and rapid clinical events. CONCLUSIONS This phenotype series offers a robust, reusable framework for measuring adoption of the Wake Up and Breathe protocol. Despite limitations in EHR data availability, it supports scalable quality measurement and retrospective analysis across ICU settings. Future work should address extending the definitions to include currently unmeasured protocol components.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.