UMMS SEPSys sepsis prediction and RESCUE deterioration risk scoring for adult inpatients

Full Title: Sepsis Early Prediction System Score (SEPSys Score) and RESCUE Score Multi-Site, Two-Intervention, Cluster-Randomized, Factorial Stepped-Wedge Pragmatic Clinical Trial at the University of Mayland Medical System

Not applicable Interventional University of Maryland, Baltimore · NCT06779617

This project will test two four-hour risk scores that alert hospital providers about adult inpatients' near-term risk of developing sepsis or sudden clinical deterioration.

Quick facts

PhaseNot applicable
Study typeInterventional
Enrollment150000 (estimated)
Ages18 Years and up
SexAll
SponsorUniversity of Maryland, Baltimore Academic / other
Locations8 sites (Baltimore, Maryland and 7 other locations)
Trial IDNCT06779617 on ClinicalTrials.gov

What this trial studies

Researchers at the University of Maryland Medical System will deploy two machine-learning risk scores — the SEPSys score for four-hour sepsis risk and the RESCUE score for four-hour deterioration risk — each producing a color-coded 1–4 risk level. The study runs on participating adult inpatient units where the scores are calculated for bedded patients and shown to clinicians during usual care. The primary outcome is the time from when a risk score becomes elevated to when vital signs (for example heart rate or blood pressure) are next measured as a marker of increased provider awareness. These tools are predictive aides meant to prompt closer monitoring and are not diagnostic of an active event.

Who should consider this trial

Good fit: Adult inpatients (age 18 or older) on participating UMMS hospital units who have SEPSys and/or RESCUE scores available are the ideal candidates for this study.

Not a fit: Patients under 18, encounters on nonparticipating units (including units that are outpatient, pediatric-only, psychiatric, or routine OB deliveries without scores) or patients without SEPSys/RESCUE scores are unlikely to receive benefit from this intervention.

Why it matters

Potential benefit: If successful, the scores could help clinicians notice at-risk patients sooner so they monitor and respond earlier, potentially preventing arrests, unplanned ICU transfers, or worsening sepsis.

How similar studies have performed: Prior early-warning and machine-learning approaches for sepsis and deterioration prediction have shown predictive accuracy in research settings but have produced mixed results on changing provider behavior or clinical outcomes in routine practice.

Eligibility criteria

Show full inclusion / exclusion criteria
Hospital Unit Inclusion Criteria:

* Units are part of a member hospital of the University of Maryland Medical System (UMMS)
* Units treat inpatient patients
* Units agree to participate in the study

General Encounter Inclusion Criteria:

* Encounter is on a participating inpatient unit
* Encounter is associated with a bedded patient in the hospital
* Encounter is associated with an adult (at least 18 years of age at time of admission)

Additional SEPSys Encounter Inclusion Criteria:

• Encounter has a SEPSys Score

Additional RESCUE Encounter Inclusion Criteria:

• Encounter has a RESCUE Score

Exclusion Criteria:

Hospital Unit Exclusion Criteria:

* Unit treats exclusively outpatients or is located in an outpatient setting
* Unit treats primarily or exclusively pediatric patients (patients under the age of 18 years old)
* Unit treats exclusively psychiatric patients

Additional SEPSys Encounter Exclusion Criteria:

* Encounter is with OB for a normal delivery and the patient has yet to give birth. This encounter can become eligible once the patient has given birth.
* Encounter is admitted with a diagnosis of sepsis
* Sepsis was already diagnosed during the encounter

This encounter will have been eligible until the patient began showing clinical signs or symptoms of sepsis or received a diagnosis of sepsis. Once they have a diagnosis or clinical signs or symptoms of sepsis, their SEPSys Score will no longer be updated. Data will continue to be collected on time of laboratory orders and treatments.

Additional RESCUE Encounter Exclusion Criteria:

• Encounter is admitted with a high RESCUE Score (RESCUE Score of 4)

Where this trial is running

Baltimore, Maryland and 7 other locations

Study contacts

How to participate

  1. Review the eligibility criteria above with your treating physician.
  2. Visit the official trial page on ClinicalTrials.gov for the most current contact information and recruitment status.
  3. Contact the listed study coordinator or principal investigator to request pre-screening. Pre-screening is free and never obligates you to enroll.
Conditions SepsisPatient DeteriorationAIArtificial IntelligenceMachine learningDecision supportClinical decision supportEarly warning systems
Last reviewed 2026-06-10 by the Find a Trial editorial team. Information on this page is for educational purposes and is not medical advice. Always consult qualified healthcare professionals about clinical trial participation.