Research projects
We have pioneered deployable critical care informatics tools that move beyond retrospective analytics to real-time surveillance, clinician-centered decision support and outcomes-focused implementation.
Our current research portfolio emphasizes:
- High-impact artificial intelligence (AI).
- Multimodal sensing, including waveforms, device-derived data and computer vision.
- Pragmatic clinical trials.
- Scalable platforms that convert data into bedside action.
- Automated severity scoring.
- Large language model (LLM)-enabled reasoning support.
All technologies undergo rigorous validation, usability testing and pragmatic evaluation to ensure measurable improvements in patient outcomes.
Predictive surveillance for medical deterioration
Focus: Real-time risk detection and prescriptive intelligence that identifies medical deterioration conditions early, often hours before overt clinical recognition, and supports evidence-informed action. These conditions include sepsis, shock and respiratory failure.
Purpose and goals: Develop, validate and implement predictive models and sniffer architectures that are accurate, explainable and operationally deployable across units and sites.
Key findings: Validated sniffer methods can achieve strong diagnostic performance and identify meaningful delays in recognition and treatment, supporting quality improvement, research and clinical rescue workflows.
Significance to patient care: Earlier recognition of sepsis or instability reduces diagnostic delay and enables timely treatment. This can help prevent avoidable ICU escalation and complications.
Select publications
Developing the surveillance algorithm for detection of failure to recognize and treat severe sepsis. Mayo Clinic Proceedings. 2015.
Predicting sepsis onset using a machine learned causal probabilistic network algorithm based on electronic health records data. Scientific Reports. 2023.
Effectiveness of automated alerting system compared to usual care for the management of sepsis. NPJ Digital Medicine. 2022.
Implementation and evaluation of sepsis surveillance and decision support in medical ICU and emergency department. The American Journal of Emergency Medicine. 2022.
Multicenter derivation and validation of an early warning score for acute respiratory failure or death in the hospital. Critical Care. 2018.
ICU population management platforms (AWARE and AMP)
Focus: Clinician-led situational awareness for teams managing multiple patients who are critically ill simultaneously.
Purpose and goals: Build and evaluate dashboard viewers that reduce cognitive burden, speed task completion and reduce omission errors while fitting real ICU workflows.
Key findings: AWARE implementation reduced preround data-gathering time and improved perceived efficiency and cognitive workload. AMP reduced time to assessment and clinician task load compared with standard EMR workflows.
Significance to patient care: Better prioritization and faster information synthesis supports safer escalation decisions and improves reliability of multidisciplinary care.
Select publications
The implementation of clinician designed, human-centered electronic medical record viewer in the ICU: A pilot step-wedge cluster randomized trial. International Journal of Medical Informatics. 2015.
Evaluation of digital health strategy to support clinician-led critically ill patient population management: A randomized crossover study. Critical Care Explorations. 2023.
Who needs clinician attention first? A qualitative study of critical care clinicians' needs that enable the prioritization of care for populations of acutely ill patients. International Journal of Medical Informatics. 2023.
Validation of computerized automatic calculation of the sequential organ failure assessment score. Critical Care Research and Practice. 2013.
Interaction time with electronic health records: A systematic review. Applied Clinical Informatics. 2021.
Ambient intelligence and multimodal sensing in the ICU (computer vision, devices, workflow)
Focus: Treating the ICU as a complex adaptive system integrating context from people, tasks and environment to build ambient, nondisruptive intelligence.
Purpose and goals: Use computer vision and sensing and workflow science to quantify care processes, improve situational awareness and enable safer AI deployment.
Significance to patient care: Supports detection of risk patterns, reduces missed deterioration signals and enables safer automation without increasing alarm fatigue.
Select publications
Are we ready for video recognition and computer vision in the intensive care unit? A survey. Applied Clinical Informatics. 2021.
Harnessing the power of technology to transform delirium severity measurement in the intensive care unit: Protocol for a prospective cohort study. JMIR Research Protocols. 2025.
Applying an agile science roadmap to integrate and evaluate ethical frameworks throughout the lifecycle and use of artificial intelligence tools in the intensive care unit. Critical Care Nursing Clinics of North America. 2025.
Automatic ARDS surveillance with chest X-ray recognition using convolutional neural networks. Journal of Critical Care. 2024.
Classification of respiratory conditions using auscultation sound. Paper presented at: 43rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society. 2021. Guadalajara, Mexico.
Clinical data infrastructure and translational platforms (near-real-time acute care data — Mac3D platform)
Focus: Building the data backbone required for real-time AI, pragmatic clinical trials, quality improvement automation and reproducible validation at scale.
Purpose and goals: Maintain high-trust, near-real-time acute care data environments and standardized datasets that support rapid bench-to-bedside cycles.
Significance to patient care: Enables faster validation, safer deployment and continuous monitoring of model performance and clinical impact.
Select publications
Medical informatics: An essential tool for health sciences research in acute care. Biomolecules & Biomedicine. 2009.
Early computerization of patient care at Mayo Clinic. Mayo Clinic Proceedings. 2016.
Automating quality metrics in the era of electronic medical records: Digital signatures for ventilator bundle compliance. BioMed Research International. 2015.
Automatic quality improvement reports in the intensive care unit: One step closer toward meaningful use. World Journal of Critical Care Medicine. 2016.
Development of a core critical care data dictionary with common data elements to characterize critical illness and injuries using a modified Delphi method. Critical Care Medicine. 2025.
Pragmatic clinical trials of AI decision support beyond the ICU (hospital-wide translation)
Focus: Moving from model performance papers to real clinical impact testing AI tools in pragmatic trial designs.
Purpose and goals: Evaluate whether AI changes care delivery, timeliness and outcomes when integrated into routine clinical operations.
Significance to patient care: Demonstrates accountable AI, which means measurable improvements in time-to-care and patient-centered outcomes.
Select publications
Effect of an artificial intelligence decision support tool on palliative care referral in hospitalized patients: a randomized clinical trial. Journal of Pain and Symptom Management. 2023.
Improving time to palliative care review with predictive modeling in an inpatient adult population: study protocol for a stepped‑wedge, pragmatic randomized controlled trial. Trials. 2021.
Multicenter derivation and validation of an early warning score for acute respiratory failure or death in the hospital. Critical Care. 2018.
Development of a core critical care data dictionary with common data elements to characterize critical illness and injuries using a modified Delphi method. Critical Care Medicine. 2025.
Improving in‑hospital patient rescue: What are studies on early warning scores missing? A scoping review. Critical Care Explorations. 2022.
Telecritical care, control tower operations and scalable acute care oversight
Focus: Scaling critical care expertise and real-time surveillance across units and sites using centralized monitoring platforms, standardized workflows and digitally enabled oversight systems. These include developing and deploying the CEDAR platform as a next-generation operational intelligence environment.
Purpose and goals: Design and evaluate centralized control tower models for ICU and hospital-wide population management; define best operational models, evaluate barriers and facilitators, and build measurement frameworks that demonstrate value.
Significance to patient care: Supports consistent rescue workflows, improves reliability of escalation pathways and strengthens system-wide ICU operations.
Select publications
A survey of tele-critical care state and needs in 2019 and 2020 conducted among the members of the Society of Critical Care Medicine. Healthcare. 2022.
Clinical impact of intraoperative electronic health record downtime on surgical patients. Journal of the American Medical Informatics Association. 2019.
Interaction time with electronic health records: A systematic review. Applied Clinical Informatics. 2021.
Information needs for the rapid response team electronic clinical tool. BMC Medical Informatics and Decision Making. 2017.
Evaluation of digital health strategy to support clinician‑led critically ill patient population management: A randomized crossover study. Critical Care Explorations. 2023.