One of the most significant – and revealing – performance indicators tracked by large, urban, high-volume healthcare facilities is the percentage of emergency department (ED) visitors who leave without being seen (LWBS).
Some of the factors impacting how patients flow through the ED include inpatient stay duration, bed occupancy, overall hospital throughput and ED efficiency. Care demand and how resources are managed and utilized can ultimately be reflected in wait times for care within the ED. Thus, a high number of patients who leave without being seen often signals that access to care issues may be prevalent and are likely multifactorial. Patients who leave without being seen have higher rates of returning to the ED within 48 hours – often with potentially worse patient outcomes that can also negatively impact health systems financially.
When Rush University Medical Center (RUMC) faced this challenge within their own walls, they set a goal to reduce their LWBS ratio. RUMC was determined to accomplish this to improve throughput and patient satisfaction – and the medical center knew leveraging their analytics capabilities to the fullest extent would help them to achieve their goal. With this in mind, the medical center leveraged the HIMSS Analytics Adoption Model for Analytics Maturity (AMAM).
Since integrating predictive models into the clinical workflow was at the core of this effort, a multidisciplinary group of ED clinicians, executives and data scientists was assembled to build them. Using predictive and prescriptive analytics, this team was able to identify individuals at high risk for leaving the ED and customize workflows to quickly intervene before the patients were likely to leave without being seen.
These predictive models were developed through a sequential process, which included the following:
- data set extraction
- machine learning model training
- model refinement through user validation
- development of a deployment framework in the electronic health record
- production deployment
Data used for the model included medications, procedures, laboratory testing, billing history, encounter history, prior ED LWBS events, and social determinant of health data that RUMC screens for and integrates into the patient record.
For workflow integration, RUMC’s Information Systems department leveraged a cognitive computing platform that allowed for the retrieval of real-time data to deploy the output of models into workflows. The model outputs were integrated into the ED dashboard and in administrative workflows, showing the probability for LWBS per patient. Workflows were customized to intervene based on the outputs that were displayed in clinical charts.
Prior to launching this initiative, RUMC’s LWBS rate was 3.8 percent. After the predictive models were implemented to enhance clinical workflows, patients identified as likely to leave the ED dropped to 1.25 percent in mere months. Clinical return on investment was also significant following implementation. RUMC’s work to become more information-aware led to a tremendous gain for the organization, and improved care.
HIMSS Analytics AMAM Stage 7 Validation
HIMSS is pleased to recognize Rush University Medical Center for achieving Stage 7 validation with the HIMSS Analytics Adoption Model for Analytics Maturity (AMAM). The AMAM is designed to measure and advance an organization’s analytics capabilities.
RUMC is the third organization worldwide to achieve this validation. The Chicago medical center also validated at Stage 7 for the HIMSS Analytics Electronic Medical Record Adoption Model (EMRAM). In addition, RUMC was honored with the 2018 HIMSS Davies Enterprise Award.
“Rush University Medical Center has dynamic analytics leadership. From the COO’s office live dashboard to front-line care in the Road Home Program for vets, Rush showed how to deliver passionate, data-driven care,” said James E. Gaston, Senior Director, Healthcare Advisory Services Group, HIMSS Analytics.
“Data and information are the new liquid gold,” said Dr. Shafiq Rab, senior vice president and chief information officer, Rush University Medical Center and the Rush System. “At Rush, we are using tools like artificial intelligence and machine learning to mine that gold and use it to improve outcomes for our patients and our communities.”
Empower Your Electronic Medical Records
Expand the realm of possibilities for improved outcomes with HIMSS Analytics maturity models:
- Adoption Model for Analytics Maturity (AMAM)
- Continuity of Care Maturity Model (CCMM)
- Digital Imaging Adoption Model (DIAM)
- Electronic Record Adoption Model (EMRAM)
- Infrastructure Adoption Model (INFRAM)
- Outpatient Electronic Medical Record Adoption Model (O-EMRAM)
Visit the HIMSS Analytics website to learn more.