The University of Rochester received a $172,785 Project Grant award from the Agency for Healthcare Research and Quality (AHRQ) under the Research on Healthcare Costs, Quality and Outcomes (CFDA 93.226) program. The award, titled "ML-ROVER: Machine Learning to Reduce Laboratory Test Overutilization," aims to develop machine learning (ML) models and a clinical decision support (CDS) tool to predict future laboratory values and reduce medically unnecessary testing in pediatric intensive care units (PICUs).
The project is a two-stage effort. In the initial R21 phase, the University of Rochester will train ML models using data from over 188,000 PICU patient encounters to forecast future lab test results (Aim 1). Concurrently, the team will identify key contextual factors to inform the design of an ML-based CDS system (Aim 2). In the subsequent R33 phase, the researchers will incorporate new predictive features to enhance the ML models (Aim 3) and design an EHR-embedded CDS tool incorporating user-centered principles (Aim 4). The CDS tool will then be piloted at a selected PICU site to measure implementation outcomes. The project seeks to establish a generalizable approach for developing data-driven translational decision support tools that can be adapted to diverse healthcare settings and reduce disparities.
The University of Rochester has also awarded a sub-grant to Cedars-Sinai Medical Center to support the "ML-ROVER" project. Cedars-Sinai will contribute specialized expertise and services, leveraging its track record of delivering biomedical research and healthcare services through federal contracts and grants.
Generated 3/4/25, 8:15 AM