Advancing scalable AI-powered clinical decision support for AUC-guided vancomycin dosing in ICU patients

Active Grant Awarded

Award ID
R03TR006291
Recipient
Awarding agency
Obligated
$154,000

Award facts

Completion date 9% complete
Place of performance
Maywood, IL 60153, USA

Award details come from USAspending.gov.

About this award

The Department of Health and Human Services National Institutes of Health National Center for Advancing Translational Sciences awarded Loyola University of Chicago Health Sciences Campus a grant of $154,000 on September 14, 2026, under the National Center for Advancing Translational Sciences program (CFDA 93.350). The award funds project R03TR006291, which aims to move AIDX, an AI-powered clinical decision support system for AUC24-guided vancomycin dosing in ICU patients, from research prototype to a validated, scalable tool. The work integrates the recipient's VANCO24 data sniffer, which estimates AUC24 from EHR data without pharmacokinetic profiles. Aim 1 trains explainable AI models, and Aim 2 constructs a modular prescriptive engine that recommends patient-specific dosing regimens. Deliverables include a public-facing demonstration tool to engage stakeholders and prepare for clinical integration. Model development uses multi-institutional ICU datasets from Loyola, UCHICAGO, UIC, Mayo Clinic, and MIMIC-IV. The project aligns with PAR-25-155. Performance runs from the award date through August 31, 2028, with place of performance in Maywood, Illinois 60153.

Description

Abstract
Vancomycin remains a cornerstone therapy for serious gram-positive infections such as mrsa, particularly in ICU settings. However, its narrow therapeutic window poses significant risks: subtherapeutic levels can lead to treatment failure and resistance, while supratherapeutic dosing increases nephrotoxicity. Recent guidelines recommend dosing based on the 24-hour area under the concentration-time curve (AUC24), but widespread adoption remains limited due to workflow barriers and reliance on costly bayesian software. This project proposes to transition aidx, an AI-powered clinical decision support system (CDSS), from a research prototype to a validated and scalable tool aligned with AUC24-based dosing. Building on KL2- supported foundational work, this R03 will integrate our new VANCO24 data sniffer, which enables automated AUC24 estimation from EHR data without pharmacokinetic profiles. Leveraging multi-institutional ICU datasets (loyola, uchicago, UIC, mayo clinic, and mimic-IV), we will train explainable AI models (aim 1) and construct a modular prescriptive engine to recommend patient-specific dosing regimens (aim 2). The final output will include a public-facing demonstration tool to engage stakeholders and prepare for clinical integration. The long-term goal is to improve antibiotic stewardship and patient outcomes in ICU settings by deploying scalable, interpretable, and generalizable AI tools. This project aligns with PAR-25-155 by addressing a clear translational science GAP and supporting the PI's transition to independent research.

Quick answers

Who received award R03TR006291?

R03TR006291 was awarded to Loyola University Of Chicago by National Center for Advancing Translational Sciences (National Institutes of Health) on September 14, 2026.

How much is award R03TR006291 worth?

$154,000 has been obligated on R03TR006291.

On GovTribe

See this award's full record

  • Every transaction and subaward
  • The recipient's other awards and teaming partners
  • Similar grant awards
  • Ask GovTribe AI about this award