June 1999
By Charles Branas and Charles ReVelle
Wilderness County paramedics are dispatched to the scene of a motorcycle crash. Upon arrival, the paramedics determine that the motorcycle's rider has a broken arm but is otherwise stable. A radio conference with the medical command physician at Wilderness County Community Hospital determines that they should immediately make the 20-minute drive to the hospital's emergency department. After safely immobilizing their patient on a wooden backboard, they depart for the hospital.
Five minutes en route, their talkative, otherwise stable patient, loses consciousness. Shallow breathing and dangerously low blood pressure mean his condition has become severe. Although Wilderness County Community Hospital provides quick access to emergency medical care by ground, it is not equipped to handle severe, life-threatening injuries. Both paramedics are aware of this.
After another short radio conference with medical command, a rendezvous with Big City Trauma Center's helicopter is established at a nearby schoolyard. The paramedics know that Big City Trauma Center is over an hour's drive from Wilderness County, and in order for them to take advantage of the helicopter's speed, they will have to reach and prepare the rendezvous site within 15 minutes, the approximate travel time for the helicopter. They are now in a race to make certain that the decision to fly their patient to the trauma center is significantly better than the hour-long drive they have chosen not to make.
Through careful but speedy driving, the paramedics arrive at the rendezvous point in time to secure a landing zone before the helicopter touches down. Their patient is successfully transferred to Big City's helicopter, and 15 minutes later he is with a neurosurgeon in Big City Trauma Center's operating room. What could have been an hour's drive was reduced to 30 minutes, saving a life and potentially preventing irreversible disability. Medical personnel working in a well-organized medical system have won the race.
Injury (also referred to as "trauma" in medical circles) is defined as any unintentional or intentional damage to the body resulting from sudden exposure to thermal, mechanical, electrical or chemical energy. Throughout the world, injuries are the leading cause of death during half the human lifespan. On a typical day in the United States, more than 170,000 men, women and children are injured severely enough to warrant medical care. More than 400 of these people die as a result of their injuries.
Unlike many other diseases, surviving a severe traumatic injury is particularly time-dependent. For a severely injured individual, the risk of death increases three-fold after one hour without the unique surgical care of a trauma canter. This is often referred to as the "golden hour." Because the time from notification to initial field assessment by paramedics may take as much as 30 minutes, many trauma systems require 30 minutes or less to get a severely injured patient from a scene of injury to a trauma center.
The situation in rural areas like Wilderness County is not atypical. Fortunately for the injured motorcyclist in this scenario, Wilderness County and Big City are part of an organized system of medical care that designates hospitals and ambulances specifically for severe traumatic injury. Big City Trauma Center, unlike Wilderness County Community Hospital, had the facility and human resources necessary to be designated a trauma center, making it an appropriate destination for a severely injured motorcyclist.
Coordinated Response
Before the concept of a "trauma system," severely injured patients were transported to the nearest hospital regardless of that hospital's ability to address their level of injury. Today, trauma systems in the United States involve coordinated response plans to ensure that severely injured individuals have appropriate and timely access to dedicated personnel at specialized trauma center hospitals. This strategy has been highly effective in reducing preventable death and disability due to injury.
Trauma centers were originally inner-city hospitals that had assumed de facto trauma center status. Unfortunately, any hospital that satisfied the necessary clinical requirements proclaimed itself a trauma center. Rarely, if ever, was a systematic assessment of its geographic relationship to ambulance facilities or its likely impact on regional demands conducted. Failure to limit the number of trauma centers based on demonstrated geographic need resulted in a "medical arms race" among trauma systems that produced organizational hypertrophy yet offered little marginal benefit to the public health.
Attempts in the early 1980s to recognize and close surplus trauma facilities did little to restrain the growth of trauma systems. As of 1993, 21 states were served by legally authorized statewide or regionally based trauma systems. The brief funding provided by the Trauma Systems Development Act, which was repealed in 1996, encouraged an additional 19 states to initiate publicly administered trauma systems of their own. Nevertheless, geographic standards and the reduction of duplicative resources remained vital to the continued existence of solvent trauma systems.
Today, designated trauma centers function at the heart of the U.S. trauma system. They remain a "safety net" for severely injured individuals despite inconsistent political and financial support. The question of exactly how many trauma centers are necessary to maintain an adequate safety net, however, is still largely unanswered. The American College of Surgeons estimates that one trauma center per million population is sufficient to address typical volumes of severely injured patients and maintain the expertise of medical providers. This estimate is largely speculative, though, and only provides vague guidance in resolving where and how many trauma centers should be located for a particular region.
Past studies of emergency medical services have addressed the location of ground ambulances and their destination hospitals almost always in separate analyses. However, emergency medical services location analysis has, to the best of our knowledge, not been used to simultaneously locate ambulances and hospitals strictly for trauma. Consequently, the incorporation of location techniques that had proven beneficial to the patient broadly in need of emergency care would also be of value in addressing the more specific needs of the trauma patient.
In formulating a new technique for trauma resource location, covering concepts were applied from previous location models. These concepts were combined and extended to create the Trauma Resource Allocation Model for Ambulances and Hospitals or TRAMAH.
TRAMAH Model
The TRAMAH simultaneously locates both trauma centers and helicopter depots and measures success in terms of the number of severely injured persons having timely access to a trauma center by either ground or air. A severely injured person that cannot be driven or flown to a trauma center within a prespecified response time standard lacks access and is, presumably, a candidate for the morgue.
In a study to test the effectiveness of TRAMAH, nearly 27,000 cases of severe injury were analyzed in the state of Maryland between 1992 and 1994. These injuries were separated into ZIP codes and a "spatial injury profile" was created showing two high-volume clusters of severe injuries, in the Baltimore and metropolitan Washington, D.C. areas.
Because trauma centers and helicopter depots have little meaning as fractions (for instance, 3 1/4 trauma centers or 5 1/2 helicopter depots), TRAMAH needed to be formulated as an integer programming model in order to site only whole numbers of trauma centers and helicopter depots. Combined with heuristic concentration, a new metaheuristic that modifies a base vertex exchange heuristic, the integer programming formulation of TRAMAH was successfully used to site as many as 10 trauma centers and 10 helicopter depots at once.
The TRAMAH was engineered with the computational flexibility to locate various numbers of helicopter depots and trauma centers as separate resources or together. It can thus be engineered to build a relatively new regional or state trauma system from a "clean slate" or to accommodate partially developed or well-developed systems that are seeking only incremental improvements to their system.
Although Maryland has a well-developed trauma system, the use of TRAMAH to analyze and relocate all existing trauma centers and helicopter depots from a clean slate extended coverage to nearly all of the state's severely injured residents. Several hundred more severely injured individuals per year would have been offered access to the trauma system within 30 minutes under TRAMAH's full reconfiguration of trauma centers and helicopter depots. Conversely, Maryland could have retained its existing level of access with one less trauma center and six fewer helicopters, a substantial savings to the state.
Nonetheless, because Maryland has one of the most advanced trauma systems in the United States, relocating all trauma centers and helicopter depots as if the state were a clean slate may not be reasonable. The TRAMAH, therefore, was also designed to simulate incremental changes. In this way, holding in place the trauma centers of the existing system and optimally relocating all existing helicopter depots with TRAMAH could offer access to hundreds more severely injured residents each year. This comes at the small price of moving helicopter resources that pose relatively few obstacles to relocation.
Although the relocation of trauma centers is more difficult than that of helicopter depots, TRAMAH could also guide health planners to the best locations for trauma centers. At the very least, Maryland could reap the financial benefits of eliminating over half of its existing helicopter depots, optimally relocating those that remain, and still maintain approximately the same level of access for its severely injured residents within 30 minutes.
Improve the Existing System
Why was TRAMAH able to improve the existing system?
Reason 1: Trauma systems have traditionally sought to maximize coverage of land area, a strategy that has led health planners to locate many trauma centers and helicopter depots at the same sites. In contrast to the land coverage objective, TRAMAH maximized coverage of demonstrated need. A greater number of severely injured individuals in need of trauma care were thus offered access to the trauma system through helicopter depots optimally located as "satellites" to one or more trauma centers.
Reason 2: The technical tools used to locate trauma system resources are often crude. Even a moderately sized region with the intent of siting a small number of resources can generate many more configuration choices than can effectively, much less optimally, be evaluated by a trauma systems administrator who is simply looking at a map for guidance. The TRAMAH can assess hundreds of thousands of possible trauma center and helicopter depot location combinations in a comparably short period of time.
Trauma systems require trauma resource investments that efficiently protect the health of the community. The race from Wilderness County to Big City can be won more often with a well-organized, well-planned trauma system. The TRAMAH is a powerful tool that can help trauma systems cost-effectively achieve the goal of citizen protection and improve the survival of their most grievously injured patients.
| Acknowledgement |
| This article was adapted from two manuscripts: "An iterative switching heuristic to site hospitals and helicopters" (in press by Branas and ReVelle for Socio-Economic Planning Sciences) and "A trauma resource allocation model for ambulances and hospitals" (in press by Branas, MacKenzie and ReVelle for Health Services Research). The TRAMAH was developed under grant R03H509326, "A spatial model of trauma facilities," from the U.S. Department of Health and Human Services Agency for Health Care Policy and Research. |
Charles Branas is a fellow in the Department of Health Policy and Research at the University of California at Berkeley, School of Public Health. He has published papers on various aspects of health care quality and is a former emergency medical services provider. He is also a member of INFORMS and the American Public Health Association Injury Control and Emergency Health Services Section. Dr. Branas can be reached via e-mail at: cbranas@socrates.berkeley.edu.
Charles ReVelle is a professor in the Department of Geography and Environmental Engineering at the Johns Hopkins University where he participates in the Program in Systems Analysis and Economics for Public Decision Making. He has studied location modeling and the subjects of applied integer programming for many years. He has also published more than 160 papers in OR journals and is the author or co-author of eight books. Dr. ReVelle can be reached via email at: revelle@jhu.edu.
OR/MS Today copyright © 1999 by the Institute for Operations Research and the Management Sciences. All rights reserved.
Lionheart Publishing, Inc.
2555 Cumberland Parkway, Suite 299, Atlanta, GA 30339 USA
Phone: 770-431-0867 | Fax: 770-432-6969
E-mail: lpi@lionhrtpub.com
URL: http://www.lionhrtpub.com
Web Site © Copyright 1999 by Lionheart Publishing, Inc. All rights reserved.