Personalized treatment recommendations for critically ill patients using causal machine learning
Official title: Using causal machine learning for personalized treatment recommendations in critically ill patients
Funded at University of Wisconsin-Madison · NIH-11447094 · Data updated 2026-11-12
Aims to tailor treatments for critically ill adults by using advanced machine learning to predict who will benefit or be harmed by a given therapy.
Quick facts
| Grant type | R01 grant |
|---|---|
| Study type | NIH-funded research |
| Funding institution | University of Wisconsin-Madison NIH-funded |
| Lab location | 1 site (Madison, United States) |
| Project start | August 1, 2021 |
| Project end | May 31, 2030 |
| Funding year | 2026 |
| Award (this year) | $815,985 |
| NIH project number | 2R01HL157262-05A1 |
| Project ID | NIH-11447094 on NIH RePORTER |
Who could benefit from this research
✓Could help
Adults (21+) hospitalized with critical illness who are candidates for therapies studied in critical care trials.
✕Less likely to help
Patients whose characteristics do not align with the model’s predictive signals may not benefit from personalized recommendations.
What this research studies
This project uses patient data from adults with critical illness to build models that estimate individual treatment effects, going beyond average trial results. It compares randomized trial data with real-world observational data to improve predictions. The goal is to help clinicians choose therapies that are more likely to help a specific patient based on their unique characteristics.
Why it matters
Potential benefit: Could lead to more effective, personalized therapies and better outcomes for critically ill patients.
How similar studies have gone: Previous work shows potential in estimating individualized treatment effects, but best-performing methods in critical care remain to be established.
Where this research is happening
-
University of Wisconsin-Madison
Madison, United States Active
Researchers
- Principal investigator: Churpek, Matthew Michael — University of Wisconsin-Madison
- Study coordinator: Churpek, Matthew Michael
Related research
Other NIH-funded projects on the same topics.
-
Personalized oxygen therapy for critically ill adults
A study tests whether customizing oxygen levels for adults on invasive ventilation can lower death rates and complications.
-
Imaging and EHR data combine to predict ICU outcomes with clinician and patient input
Aims to build a computer model that uses chest X-ray images, doctor notes, and electronic health record data to predict outcomes for ICU patients, with input from clinicians, patients’ families, and ethics experts.
-
Teleconsultation to Improve Safety for Critically Ill Infants and Children Before hospital arrival
A national study tests whether remote pediatric expert advice during ambulance transport reduces harm to critically ill babies and kids.
-
Patient-ventilator Asynchrony in Critically Ill Children
This grant explores how often and why mechanically ventilated children have mismatches between their own breathing efforts and the ventilator, with the goal of making ventilation safer and more comfortable for kids.
-
Patient-ventilator asynchrony in critically ill children
A study to understand when ventilators don’t perfectly match a child’s breathing needs and how this affects recovery.
-
BrighT STAR SIRIUS: a toolkit to cut unnecessary blood cultures in sick kids
A multi-site study is testing a toolkit to reduce unnecessary blood culture tests in critically ill children to prevent unnecessary treatments and harms.
About this research
- This is an active NIH-funded research project — typically early-stage science, not a clinical trial accepting patient enrollment.
- Some NIH-funded labs run parallel clinical studies or seek volunteers for related work. To check, contact the principal investigator or institution listed above.
- For full project details, budget, and progress reports, visit the official NIH RePORTER page.