Presentation
System Engineering Approach to Understand Sepsis Care Workflow during Emergency Department Admissions
DescriptionSepsis is characterized by organ dysfunction resulting from infection, with no reliable single objective test and presenting difficulty in diagnosis. There is a national challenge of sepsis detection and treatment; sepsis affects more than 1.7 million Americans annually, is a leading cause of death in US hospitals, and is the most expensive ailment treated in US hospitals, costing more than $20 billion a year. Most of these cases (>80%) are diagnosed on admission and receive initial care in emergency departments. Furthermore, every hour of delayed diagnosis (and treatment) critically increases mortality rates. National strategies for early identification have been ongoing for over a decade, and some data-based screening tools are now available for early detection.
The use of well-designed AI-based predictive analytics tools to support point-of-care clinical decision making (AI-CDSS technology) is a potential solution to help address sepsis treatment and care. Although various AI-based CDSS have been developed, a critical bottleneck, however, is in intuitive design and integration of such decision-support systems with clinical workflows. Poor integration with clinical workflows has led to limited improvements in outcomes because of ineffective work design and resistance from care-provider teams to trust/adhere to such predictive diagnostics tools, thereby increasing their mental workload.
Clinicians and administrators at Prisma Health have developed and implemented best practice strategies to improve sepsis care by developing standardized care modules and CDSS in their electronic health record (EHR) system. However, the utilization of this sepsis tool has been limited due to friction of incorporating the tools into the current clinical workflows of both physicians and nurses. We will take a systems-approach to understand the impacts of the technology with the current clinician workflow, patient care, and hospital infrastructure. The Systems Engineering Initiative for Patient Safety (SEIPS) model is leveraged to complete a novel mapping of the clinical work system components of sepsis treatment and care in the emergency department, as it relates to people (physicians, nurses, and patients); tools and technology (AI-CDSS and EHR system); care tasks; internal and external environments (physical environments and regulation/ policies); and organization (hospital leadership). The system component mapping is translated into visualizations to clearly identify the interdependencies among the work.
Our multidisciplinary team of clinicians, engineers, and data scientists will use the SEIPS model applied to sepsis on emergency department admissions to understand and evaluate the quantifiable effects and workflow integration of the current AI-CDSS technology to improve patient safety.
The use of well-designed AI-based predictive analytics tools to support point-of-care clinical decision making (AI-CDSS technology) is a potential solution to help address sepsis treatment and care. Although various AI-based CDSS have been developed, a critical bottleneck, however, is in intuitive design and integration of such decision-support systems with clinical workflows. Poor integration with clinical workflows has led to limited improvements in outcomes because of ineffective work design and resistance from care-provider teams to trust/adhere to such predictive diagnostics tools, thereby increasing their mental workload.
Clinicians and administrators at Prisma Health have developed and implemented best practice strategies to improve sepsis care by developing standardized care modules and CDSS in their electronic health record (EHR) system. However, the utilization of this sepsis tool has been limited due to friction of incorporating the tools into the current clinical workflows of both physicians and nurses. We will take a systems-approach to understand the impacts of the technology with the current clinician workflow, patient care, and hospital infrastructure. The Systems Engineering Initiative for Patient Safety (SEIPS) model is leveraged to complete a novel mapping of the clinical work system components of sepsis treatment and care in the emergency department, as it relates to people (physicians, nurses, and patients); tools and technology (AI-CDSS and EHR system); care tasks; internal and external environments (physical environments and regulation/ policies); and organization (hospital leadership). The system component mapping is translated into visualizations to clearly identify the interdependencies among the work.
Our multidisciplinary team of clinicians, engineers, and data scientists will use the SEIPS model applied to sepsis on emergency department admissions to understand and evaluate the quantifiable effects and workflow integration of the current AI-CDSS technology to improve patient safety.
Event Type
Oral Presentations
TimeMonday, March 252:10pm - 2:30pm CDT
LocationSalon A-1
Hospital Environments