The Emergency Care Systems Lab engages in projects ranging from developing predictive analytics and clinical decision support tools to conducting community-engaged research and quality improvement initiatives. We partner with local and national organizations to ensure our innovations are grounded in real-world applications and provide measurable impacts in emergency care.
To learn more about our work or express interest in partnering on research projects, please reach out to us at ECSL@g-groups.wisc.edu.
Our active and completed projects are listed below.
Active Projects
This is an accordion element with a series of buttons that open and close related content panels.
Multi-site Trial to Improve Dementia Care in the Emergency Department (ED-LEAD)
This $55-million-dollar study involves an 80-site pragmatic clinical trial testing three different interventions to improve emergency care for patients with dementia who are discharged home. One of the three interventions is the community paramedic-led transition intervention (CPTI) previously developed and tested here by Dr. Shah and his team.
The project’s specific aims are to: 1) optimize a concurrently run emergency care redesign, nurse-led telephonic care, and community paramedic-led transitions intervention in PLWD for feasibility, fidelity and usability in two EDs; 2) study the effectiveness of these three interventions, alone and in combination, for PLWD with serious illness in a cluster-randomized multifactorial trial embedded within 80 EDs on: ED revisits, hospitalizations, and healthy days at home following the index ED visit; and 3) determine site, provider, patient, and care partner-level characteristics within a diverse population associated with variation in implementation of each intervention.
Building on this evidence, the overarching goal of EDs LEading the Transformation of Alzheimer’s and Dementia Care (ED- LEAD) is to turn an ED visit from a crisis into an opportunity to improve the well-being of PLWD and their care partners.
Principle Investigator: Manish N. Shah, MD, MPH
ECSL Co-Investigator: Brian Patterson, MD, MPH
Funding source: National Institute on Aging (NIA), 1U19AG078105
Multi-site Implementation of Clinical Decision Support to Prevent Future Falls (R18)
This project focuses on reducing fall risk in older adults after emergency department (ED) visits, which is a leading cause of injury in this population. Leveraging advanced technology, we use an innovative Clinical Decision Support (CDS) system that integrates with electronic health records (EHR) to identify patients at high risk of falling.
The CDS analyzes patient data during the ED visit and, if a high risk is detected, alerts the clinician and recommends a referral to a specialized falls prevention clinic. The system is designed to work seamlessly within existing workflows and adjusts in real time based on clinic availability to prioritize high-risk patients.
Currently being tested at three UW Health EDs, the project seeks to improve referral rates, enhance preventive care, and reduce future fall-related injuries — paving the way for broader adoption across health care systems.
Principle Investigator: Brian Patterson, MD, MPH
ECSL Co-Investigators: Hanna Barton, PhD, Dough Wiegmann, PhD
Funding source: Agency for Healthcare Research and Quality (AHRQ), R18HS027735
Publications: View a list of publications
Human Factors & Predictive Analytics to Improve Patient Flow
In this project, we are implementing an electronic health record (EHR)-integrated machine learning model to predict patient admission to an inpatient unit from the emergency department. The interdisciplinary team — blending operations and research — validated a vendor-trained model, mapped the patient admission process, and explored how our model’s output predictions could support clinical decision-making.
Principle Investigator: Brian Patterson, MD, MPH
ECSL Co-Investigator: Hanna Barton, PhD
Funding source: UW Institute for Clinical and Translational Research (ICTR) Learning Health System, funded through the National Institute of Health Clinical and Translational Science Award, UL1TR002737
Human Factors & Simulation to Inform East Madison Hospital Redesign
In partnership with UW Health, this project provides advanced simulation and human factors support for redesigning the East Madison Hospital ED, while generating evidence-based toolkits to inform future hospital and ED designs.
Principle Investigators: Hanna Barton, PhD, Bob Batt, PhD, Hani Kuttab, MD
Informatics to Improve Prehospital Care
Emergency medical services (EMS) traditionally deliver life-saving medical care to the community. As part of its community commitment, UW Health is developing operational partnerships to expand this scope, delivering innovative services to both EMS agencies and local residents by analyzing clinical data generated during EMS calls. A major barrier to meeting the needs of operational stakeholders and enabling researchers to create generalized knowledge, however, is that EMS data is collected and stored separately from hospital data.
Through ongoing operational initiatives at UW Health, an information exchange linking EMS and hospital data is being created. In addition to simply linking hospital and EMS data, ECSL aims to create a secure, cloud-based computing environment with linked data accessible to both researchers and operational stakeholders, eliminating the need to work across disparate information systems and breaking down silos between research and quality improvement activities. This will create an unprecedented resource that enables community-centered research and health disparities research.
Principle Investigator: Michael Spigner, MD, NRP
Completed Projects
This is an accordion element with a series of buttons that open and close related content panels.
Developing Machine Learning to Predict Future Falls of Older Adults (K08)
This project focused on identifying and reducing the risk of falls in older adults after they are discharged from the emergency department (ED). Falls are the leading cause of traumatic injuries in this population, and many older adults who visit the ED are at a heightened risk for future falls. The project used innovative data-driven approaches to address this issue by developing and testing a predictive algorithm and Clinical Decision Support (CDS) tool that uses electronic health records (EHR) data.
The project included three main phases: 1) Data Extraction and Risk Identification: Using a combination of programmatic data extraction and natural language processing (NLP), the system identifies key risk factors for falls from patient data in the EHR. This phase compared EHR-based screening with in-person screenings to validate accuracy. 2) Predictive Algorithm Development: A machine learning-based algorithm was developed to predict which ED patients are at high risk of a significant fall within six months after their visit. The goal is to use EHR data to accurately stratify patients based on their fall risk. 3) CDS Intervention: The final phase involved creating and piloting a CDS tool that alerts ED clinicians when a patient is at high risk for falls. The tool assists clinicians by recommending fall-prevention interventions, such as medication adjustments and referrals to a falls prevention clinic. The CDS is designed to integrate seamlessly into the clinical workflow.
By improving the identification and management of fall risk in older adults, this project aimed to prevent future falls and improve outcomes for ED patients.
Principle Investigator: Brian Patterson, MD, MPH
Funding source: Agency for Healthcare Research and Quality (AHRQ), K08HS024558
Publications: View a list of publications
Human Factors to Improve Older Adult ED Care Transitions (Patient Safety Learning Lab)
The Patient Safety Learning Lab aimed to ensure safer emergency department (ED) care for older adults by developing a system to detect and recover from errors, anticipate safety risks, and improve communication and coordination. A transdisciplinary team—engineers, researchers, nurses, physicians, and pharmacists—partnered with academic and community EDs to map older patients’ care journeys and identify key barriers and opportunities for improvement.
The lab’s goals were to:
- Design and evaluate a system of care, called the patient safety passport, to support older adults after an ED visit; and
- Establish a transdisciplinary Patient Safety Learning Lab (PSLL) to engineer safer care journeys for vulnerable patients, including older adults.
Their research led to the development of a patient safety passport, inspired by a travel passport, which is checked at each transition point to highlight and address safety needs. The passport consists of three interventions:
- ED Discharge Intervention: Targets the transition home from the ED with guidance on medication self-management, care plans, follow-up, and red flags.
- ED-to-SNF Antibiotic Stewardship: Automatically identifies skilled nursing facility (SNF) residents who present to the ED and triggers communication steps to improve antibiotic use for suspected urinary tract infection, including feedback loops with SNF nurse practitioners regarding culture results and a mechanism for the nurse practitioner to follow up with the health system regarding the outcome of the antibiotic prescribed.
- Patient Journey Map: Helps patients and caregivers understand their ED experience, as broken three stages: intake and triage, assessment and diagnosis, and discharge.
Principle Investigators: Pascale Carayon, PhD, Maureen Smith, MD, MPH, PhD
ECSL Co-Investigators: Brian Patterson, MD, MPH, Manish N. Shah, MD, MPH
Funding source: Agency for Healthcare Research and Quality (AHRQ), R18HS26624
Publications: View a list of publications
Wisconsin Real-time Emergency Department COVID-19 Surveillance and Responsive Training (WIRED-RT)
The WIRED-RT project, funded by the Wisconsin Partnership Program, developed and tested an advanced surveillance and training system to support emergency departments during infectious disease surges, including COVID-19.
Led by Dr. Manish Shah and a multidisciplinary team from UW Health, the University of Wisconsin, and Marshfield Clinic Research Institute, the initiative harnessed electronic health record data and predictive analytics to identify early signals of respiratory virus outbreaks. This enabled just-in-time, simulation-based training for frontline clinicians, supporting timely and coordinated responses. The project culminated in the creation of scalable training materials and contributed to peer-reviewed dissemination, including an AMIA publication on forecasting respiratory virus flares across multiple EDs.
Although originally focused on COVID-19, the framework is adaptable to a broad range of public health emergencies. The effort highlights the critical role of informatics and provider readiness in emergency preparedness.
Principle Investigator: Manish N. Shah, MD, MPH
ECSL Co-Investigators: Brian Patterson, MD, MPH, Michael Pulia, MD, PhD, Justin Boutilier, PhD
Funding source: Agency for Healthcare Research and Quality (AHRQ), R18HS027735
Publications: View a list of publications