Computational Modeling of Emergency Medical Services Aaron Bair

Computational Modeling of Emergency Medical Services Aaron Bair, MD Emergency Medicine UC Davis Medical Center

Overview • Background • Current status • The future

Background • Multiple contributing factors make this necessary and possible – Crisis-level overcrowding problems have led to increased interest in studying and promoting ED efficiency – Bioterror and disaster preparedness (surge) – Computer simulation has been used successfully in other industries for decades (manufacturing) – Hardware and software advances

What is a “model”? • Epidemiological, Statistical and CS definitions – Overlapping considerations • Discrete Event Simulation – Ability to model multiple discontinuous events with probabilistic input

Limitations • GIGO applies! – Limited by the accuracy of input data – Limited by understanding of complex processes – Limited by interpretation of complex output

EDSIM 2. 12© • • • 12, 500+ hierarchical computational modules Representative model of UCDMC ED Stochastic inputs for laboratory turn around times 3, 000 representative patients drawn from UCDMC ED cohort Patient path step approach Full activity pre emption

The Team • • Aaron Bair, MD – Emergency Medicine Lloyd Connelly, MD, Ph. D – Model engineer Beth Morris, MPH – Project Manager, Data Manager Alex Tsodikov, Ph. D Statistician Lauri Dobbs, Ph. D – Engineer, LLNL Michael Johnson, Ph. D – Engineer, Sandia Nathaniel Hupert, MD, MPH – Modeling and Outcomes research, Cornell • Nathan Kuppermann, MD, MPH – Research Mentor

EDSIM recent applications • Triage strategy analysis – Standard v. Acuity Ratio Triage 1 • Nursing shortage: RN allocation strategy analysis – Partial v. Complete area closure • Quality of care – Implications of crowding: Resource saturation impact on cardiac chest pain patients 2 1. 2. Connelly LG, Bair AE. Discrete Event Simulation of Emergency Department Activity: A Platform for System Level Operations Research. Acad Emerg Med. 2004; 11: 1177 -1185. Connelly LG, Bair AE. Computer Simulation and Observational Study of the Cardiac Chest Pain Patient in a Variably Overcrowded ED. Acad Emerg Med. In Press.

Advantages of modeling • Detailed model can be used for more mundane work flow efficiency projects • Representative model can be used as “pretrial” for extraordinary what-if scenarios – Scenarios that will probably never be prospectively studied

Next steps EDSIM Validate and merge Goal: A generalized hospital model to study both routine work flow and crisis optimization (disaster response)

The Bio. Net model The program seeks to improve the ability of a major urban area in the United States to manage the consequences of a biological attack on its population and critical infrastructure by integrating and enhancing currently disparate military and civilian detection and characterization capabilities. Combined Hospital Simulator Modification size and resources

A vision of the future • Expand collaborative relationships to create a model that can be used to analyze and optimize patient flow under variable circumstances – UCDMC: Emergency Services model (EDSIM) – Cornell University: Hospital based services model (AHRQ project) – Oregon Health Sciences: Center for Policy Research in EM – Sandia National Laboratories: Bio. Net project and regional model (http: //bionet. calit 2. net/project. php) (NDA in place) – Lawrence Livermore National Laboratories: model validation (HS grant funded) – Look Ahead Decisions Inc: Optimization project (NLM grant decision pending) – NCEMI – Project Sentinel: azyxxi (Washingon D. C. )

More thoughts on the future • Optimization research • Dual supervised Ph. D grad student -Funding source for training: DHS/Sandia? HRSA? • UC Davis EM researcher role? – Non-clinical funding • Grants? – Expansion from prior training grants? • Institutional support?

Conclusions Model uses: Preparation, Policy and Administration • Computer modeling of complex and variable systems is increasingly possible • Modeling can lead to better understanding of flow (bottleneck identification) and resource optimization strategies • Particularly valuable for rare scenario analysis and preparedness (disaster response)