How common sepsis treatments help different patients
The impact of clinical interventions for sepsis in routine care and among detailed patient subgroups: A novel approach for causal effect estimation in electronic health record data
['FUNDING_OTHER'] · STANFORD UNIVERSITY · NIH-11143932
This project uses hospital medical records and machine learning to find which routine sepsis treatments help which kinds of patients.
Quick facts
| Phase | ['FUNDING_OTHER'] |
|---|---|
| Study type | Nih_funding |
| Sex | All |
| Sponsor | STANFORD UNIVERSITY (nih funded) |
| Locations | 1 site (STANFORD, UNITED STATES) |
| Trial ID | NIH-11143932 on ClinicalTrials.gov |
What this research studies
From your perspective, researchers will analyze large electronic health records from hospitals to see how people with sepsis did after common treatments. They will combine machine learning with a method called regression discontinuity to compare patients who were just on different sides of typical treatment decision cutoffs, which can act like a natural experiment. Because this work uses existing clinical data rather than enrolling people in a trial, there is no new treatment or extra procedures for patients. The aim is to generate reliable, patient-specific information about which standard sepsis interventions are likely to help different subgroups.
Who could benefit from this research
Good fit: People who were treated for sepsis in hospitals whose electronic health records are included in the study, especially those with clear clinical measurements around treatment decisions, would be the individuals whose data inform the work.
Not a fit: Patients treated outside the contributing hospital systems, those without detailed chart data, or those with very rare sepsis causes may not be represented and thus may not see direct benefit.
Why it matters
Potential benefit: Could help doctors tailor standard sepsis treatments to the patients most likely to benefit, potentially reducing deaths and harms.
How similar studies have performed: Regression discontinuity has produced credible causal estimates in some healthcare settings, but pairing it with machine learning in large EHR datasets for sepsis is a novel approach.
Where this research is happening
STANFORD, UNITED STATES
- STANFORD UNIVERSITY — STANFORD, UNITED STATES (ACTIVE)
Researchers
- Principal investigator: GELDSETZER, PASCAL — STANFORD UNIVERSITY
- Study coordinator: GELDSETZER, PASCAL
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 below.