From Bioterrorism to Natural Disasters
Background:
Researchers at Cornell’s Weill Medical College Department of Public Health led by Nathaniel Hupert, MD, MPH, were some of the first in the country to bring the tools of computer modeling to bear on the complex planning and operational issues of public health and hospital response to mass casualty events including bioterrorism.
Over the past year, Dr. Hupert has teamed with Professor John Muckstadt and colleagues at the Cornell University School of Operations Research and Industrial Engineering to develop a series of models focusing on hospital surge capacity and the complex linkages between pre-hospital care (e.g., triage and/or prophylaxis campaigns) and regional hospital capabilities.
Challenge:
Emergency planners, service providers, and hospital managers contend with massive data flows in an environment that has the opportunity to more effectively employ technology to manage resources. Emergency surge capacity, even in the largest hospitals of our major cities, typically is managed with manual bed counts (sometimes augmented by data from the nightly automated census from the Emergency Department registry log) and phone calls. Given the current threat of intentional mass casualty events in addition to natural catastrophes, the public health system can reap greater benefits from operations research through the application of multiple modeling and planning tools.
Solution:
The Bioterrorism and Mass Casualty Response Logistics Program is designed to design, construct, test, and deploy a logistics management and decision support system that consists of three principle components: planning, response and simulation.
Planning System
Develop an environment for establishing the required quantities and disposition of various medical resources as well as rules for allocating their use in time of need.
These include: physical resources such as hospitals, beds, ICUs, emergency departments, operating rooms, doctors, nurses and other health professionals, other equipment and supplies necessary for emergency and ongoing medical response; transportation assets such as ambulances and other EMS units; command and control mechanisms that ensure the maximum effective use of these physical assets; and an identification and information system including data required to support the system.
Response System
Develop a real-time system that would be employed to maximize response logistics resources once an emergency occurs.
This system includes acquisition of state-of-the-system data in real time; evaluation of alternative actions that could be undertaken by personnel in different parts of the response system (hospital administrators, political leaders, police, emergency management personnel, regional governmental and health system managers, etc., in real time; dissemination of system status and projected consequences of alternative courses of action to appropriate personnel in real time; and a command and control infrastructure that can ensure system performance is maximized.
Simulation & Training Environment
Develop a simulation environment for constructing and assessing the effectiveness of alternative logistic response system designs.
These include physical elements, information systems, command and control architectures, management rules, etc. and to train personnel to insure that all management levels of emergency response logistics systems understand and can execute their tasks in a maximally effective manner.
Results:
While Hupert’s and Muckstadt’s prototype models are probabilistic in nature and, hence, gives stochastic estimates of various resource levels needed throughout the planning horizon, this type of dynamic, probabilistic information can be used to augment the information contained in standard reporting from hospitals (e.g., that contained in the NY State Hospital Emergency Response Data System (HERDS) database) of the number of Emergency Department or other beds.
In the future, Dr. Hupert and Professor Muckstadt will expand these models and create other stochastic, time-dependent models. For example, they will incorporate logistical issues of transporting patients throughout the effected areas to the models. This involves transporting casualties to hospitals by various means (ambulances, buses, automobiles, etc.) and evacuating and moving patients currently within hospitals to alternative locations.