Chapter 8: Triage

Chapter 8: TRIAGE

In contrast to most routine emergencies, efficient response in disasters requires procedures for triage and casualty distribution. (Courtesy of California Office of Emergency Services, Sacramento, California.)

One aspect of resource management that deserves special attention is the distribution of resources for medical care. This chapter will address the concepts of triage and how they can be applied to affect medical resource management. The more common difficulties in disaster triage and some suggested solutions are discussed.


WHAT IS TRIAGE?

Traditionally, triage has been called the keystone to mass casualty management (Bowers, 1960:59). Triage comes from the French verb, trier, which means "to sort." It evolved, perhaps as early as Napoleon's time, as a technique for assigning priorities for treatment of the injured when resources were limited. The basic concept was to do the greatest good for the greatest number of casualties. Generally, attention is given first to those with the most urgent conditions and to those who are the most salvageable (Rund, 19813; Silverstein, 1984:8). The technique is considered by many to be essential for good disaster medical care (Spirgi, 1979:25; FEMA, 1983e:108; Cowley~ 1982; Burkle, 1984:45).

Doing the greatest good for the greatest number of disaster casualties does, however, involve more than just deciding who gets treated first. It also requires that use of all of the available treatment resources is maximized. That is, that the casualties are distributed rationally among the various hospitals and other medical treatment facilities (Silverstein, 1984:8, 44). Therefore, the definition of triage to be used in this text includes the organized evaluation of all disaster casualties to establish treatment and transport priorities. In addition, it involves the process by which casualties are rationally distributed among the available treatment facilities. Typically, management of triage is a systems problem requiring inter-organizational coordination and flow of information.

There are three major reasons why triage is beneficial in the disaster response:

1. Triage separates out those who need rapid medical care to save life or limb.
2. By separating out the minor injuries, triage reduces the urgent burden on medical facilities and organizations. On average, only 10-15% of disaster casualties are serious enough to require over-night hospitalization.
3. By providing for the equitable and rational distribution of casualties among the available hospitals, triage reduces the burden on each to a manageable level, often even to "non-disaster" levels.

TRIAGE PROBLEMS IN DISASTERS

Observations in disasters have revealed problems with triage. The most comprehensive data collected to date are those from the Disaster Research Center obtained as part of a study of emergency medical services (EMS) in 29 major U.S. disasters occurring in the 1970's (Quarantelli, 1983; Golec, 1977). Because no similar studies have been carried out since then, it is difficult to determine to what extent these problems have been ameliorated by modem improvements in EMS systems and disaster planning. However, there is evidence to suggest that at least some of these problems continue to occur.

When Disaster Research Center investigators carried out these studies, they found an interesting discrepancy. In 55% of the cases studied, responders claimed that triage was carried out. However, the researchers found that the word "triage" was used in a loose fashion to describe almost any handling of the victims by emergency personnel. Sometimes the presence of uniformed medical personnel seemed to suggest to onlookers or other responders that triage was being carried out even when it was not. But if the term was used to describe appropriate assessment and sorting of all casualties according to the seriousness of their injuries, then little triage actually occurred (Quarantelli, 1983:69; Tierney, 1977:154).

Furthermore, in quite a number of disasters, casualties were not distributed among the available hospitals in a rational or efficient manner. Instead, the vast bulk of them ended up at the closest hospital, while other hospitals received no casualties at all. A variant of this pattern was where one hospital in the community was thought to give superior emergency care to critical casualties, or where it was more familiar to those providing the transportation. Such might be the case if one facility was renowned as the local "trauma center." In that event, the majority of victims sometimes ended up there (a pattern also observed in a number of previous disaster case studies) Quarantelli, 1983:73; Golec, 1977; Rosow, 1977:166; Cohen, 1982a:19; Mileti, 1975:84; Williams, 1956:659; Neff, 1977:183).

In 75% of cases studied, a majority of the casualties were sent to the closest hospital. In 46% of the cases, more than three-fourths of the casualties were sent to the nearest hospital. Only in about half of the disasters did a simple majority of the hospitals in the area receive even one casualty. The pattern is illustrated by the figures in Table 8-1.


Table 8-1. Hospital Distribution of Disaster Casualties

(Adapted from: Quarantelli EL: Delivery of emergency medical services in disasters: Assumptions and realities, Irvington, New York, 1983, p. 88.)

Another perspective on the situation is given by Table 8-2 (Golec, 1977:171). (Note the percentage treated in one hospital). The hospitals not receiving patients had an average of 20% of their beds vacant Quarantelli, 1983:79).

Not only did one hospital receive the largest number of casualties, but also those most seriously injured. In one disaster, for example, 40 out of 51 casu alties were sent to one hospital which admitted 30 of them (28 in serious condition). The remaining 11 victims were taken to four other area hospitals. Not one of these 11 had injuries considered serious enough to require staying in the hospital. The pattern was sin-War for the casualties that were dead on arrival ( Quarantelli, 1983:81). Even in those communities with only one hospital, a large community with a number of hospitals was usually located within 15 to 30 miles (Golec, 1977:172).

In considering the distribution patterns of disaster casualties, it should be noted that the optimal pattern does not necessarily mean that every hospital receives an equal number. In addition to hospital size and emergency department capacity, a facility's ability to take care of specialized cases (e.g., trauma) may affect the number of cases it can effectively handle. It could be argued that a trauma victim may be able to receive better care in a crowded trauma center than a less burdened but also less experienced facility. However, there is a lack of data on casualty severity versus receiving hospital capability in disasters, the level of case rendered, or the resulting mortality or morbidity. Therefore, the degree that overcrowding versus experience affects patient outcome is yet to be determined.

In some disasters, it has been claimed that adequate care was given, even though hospitals received a disproportionate share of casualties (KC Health Dept, 1981:13, 16, 19; Ross, 1982:65; Moore, 1958:28; Lewis, 1980:863). Nevertheless, reasonable balanced distribution of disaster casualties and especially the use of all hospitals to the levels of their capabilities, seems to be a reasonable measure of optimal medical resource use.

Despite the fact that the incidence and quality of triage has not been subjected to rigorous study since Quarantelli's research in the 1970's, anecdotal reports from recent disasters have revealed that under the right circumstances rather good patterns of casualty distribution to hospitals are possible. In some cases, use has been made of non-hospital treatment facilities for minor injuries, and casualties have been reasonably distributed among area hospitals.

Table 8-2. Distribution of Disaster Casualties

(click to enlarge)
[Deleted from this table are 3 communities with only 1 hospital]
(Adapted from Golec JA, Gurney PJ: The problem of needs assessment in the delivery of EMS, Mass Emergencies, 2:169-77, 1977.)


EXAMPLE: High-rise Fire, MGM Grand Hotel, Las Vegas, Nevada November 21, 1980. The disaster plan called for the use of the Convention Center as a secondary triage and refuge center. This facility, remaining from earlier civil defense planning, was equipped as an acute care hospital and contained 200 beds and 100 additional cots. It was staffed primarily by Red Cross and other volunteer staff Although it was not used as a hospital in this disaster, 1,700 minimally injured and displaced per-sons out of the 6,000 hotel guests were triaged to the center, many of these by bus. Of 769 injured survivors, 150 received treatment by medical teams at that location. Others were transported for treatment to the following facilities:

• Southern Nevada Hospital-104
• Desert Springs Hospital-161
• Sunrise Hospital-211
• Valley Hospital-143 (Buerk, 1982:641; Morris, 1981:20).

EXAMPLE: DC-9 Airline Crash, Stapleton International Air-port, Denver, Colorado, November 16, 1987. Nine of the area hospitals usually accept emergency ambulance patients on a daily basis. In this disaster, 10 hospitals received the following numbers of the 56 injured survivors and 5 injured rescuers:

• Denver General Hospital (a level I trauma center)-3 immediate, 24 minor;
• University Hospital (a level I trauma center)-6 immediate, 2 minor;
• St. Anthony, Central (a level I trauma center)-3 critical, 1 delayed, 2 minor;
• Swedish Hospital (a level II trauma center)-l immediate, 1 delayed, 1 minor;
• Presbyterian Aurora Hospital-2 delayed;
• St. Joseph Hospital-1 immediate, 1 delayed, 3 minor;
• St. Luke Hospital-2 delayed;
• Porter Memorial Hospital-1 delayed, 2 minor;
• Fitzimons Army Hospital-1 delayed, 1 minor;
• Rose Medical Center-2 delayed, 1 minor (Dinerman, 1988).

It should be noted, however, that in each one of these cases, the disaster covered a rather small geographic area. Adequate triage and casualty distribution is more difficult to achieve in disasters such as tornadoes, floods, hurricanes, and earthquakes, that cause injury and destruction over a wide area.

As shown in Table 8-3, observations in a number of cases reveal that communities did not take full advantage of all their available hospital resources in disasters.

Table 8-3. Distribution of Casualties

(click to enlarge)

(click to enlarge)

CAUSES OF TRIAGE PROBLEMS

Responders from Outside the Local EMS System
Non-ambulance Transport of Casualties
One of the difficulties that faces emergency medical services (EMS) systems trying to carry out triage is that many injured casualties reach the hospital outside the EMS system Quarantelli, 1983:63,70; Tierney, 1977:155; Mileti, 1975:84; Golec, 1977:175; Seismic Safety Comm, 1983:83; Scanlon, 1988:6). Table 8-4 indicates the means of arrival of the first casualties at 75 hospitals where the method of transport could be determined.

Table 8-4. Means of Initial Disaster Casualty Arrival at the Hospital

(click to enlarge)
(Adapted from: Quarantelli EL: Delivery of emergency medical service disasters: Assumptions and realities, Irvington, New York, 1983, p. 70'.)

While these figures indicate the mode of arrival of the initial casualties, overall, less than half arrived by properly equipped ambulance or rescue vehicle. The researchers noted a strong tendency for police officers to load victims into whatever vehicles were handy and send them off to the hospital. In one disaster, for example, police loaded 26 injured persons into three non-ambulance vehicles, and these were the first to arrive at the hospital (Quaran-telli, 1983:70).

Some disaster plans call for a delay in evacuation of victims from the scene. This is so they can be triaged for orderly and rational field stabilization and transport. Other plans call for the use of field first-aid stations to alleviate the burden on hospitals. However, these plans do not always take into consideration the perceptions and motivations of the victims or the public, which may be different from those of the planners. Failure to do so results in plans which may look good on paper, but which do not correspond to reality.

Often the public's perception of good emergency medical care is transportation to the hospital as quickly as possible (Drabek, 1968:148; Quarantelli, 1983:72, 110; Quarantelli, 1970a:383; Wright, 1976:27). If medical care and transportation are not furnished promptly by official emergency organizations, victims do not usually sit idly by and await its arrival. Instead, they get themselves to the hospital by the most expedient means available. Often, they will go to the nearest hospital, the one with which they are most familiar, or the one in which they have the greatest trust. Field disaster first-aid stations are often bypassed, either because their location is unknown, or because for many people "first aid" is seen as an inferior level of medical care. This pattern has been noted most particularly in diffuse, widespread disasters such as tornadoes and earthquakes (Wright, 1976:27; Quarantelli, 1970a:384; Dynes, 1974:30; Quarantelli, 1983:21,64; Raker, 1956:23; Drabek, 1986:139,170; Adams, 1981b:17,30,57; Worth, 1977:161).


EXAMPLE: Earthquake, Coalinga, California, May 2, 1983. Only 7 of the 38 casualties arriving at the Coalinga District Hospital in the first hour came by ambulance. The rest came by private car or on foot. The most seriously injured victim arrived in the vehicle of a local physician. Another local doctor, who was responsible for the city's disaster medical response, established triage site in the devastated downtown area. All patients from the downtown area, however, went directly to the hospital, bypassing the triage site (Seismic Safety Comm, 1983:83; Kallsen, 1983:25).


EXAMPLE: Tornado, Edmonton, Alberta, Canada, July 31, 1987. Out of more than 300 injured victims, 30% were transported to the hospital by a family member, 20% were taken by a stranger, 18% arrived by bus, and 16% were conveyed by ambulance (Scanlon, 1988).

This rapid transport to nearby hospitals by non-ambulance vehicles contributes to two problems seen frequently in disaster situations: 1) Casualties with relatively minor injuries arrive (often unannounced) before those with serious conditions. The result is that when the more serious victims arrive, the hospital emergency department is already inundated and its beds occupied. 2) Casualties arrive at the hospital without having been triaged or having received stabilizing first aid Quarantelli, 1983:73).

Involvement of Non-local Responders
Another factor that has contributed to the lack of organized triage and casualty distribution has been the number of responders from non-local organizations and those not under control of the local EMS system. This has been true especially in larger disasters and those occurring in urban areas (Quarantelli ' 1983:68,71; Morris, 1982a:65; Seismic Safety Comm, 1983:81,86. The increased use of helicopters for medical transport seems to contributed to this trend (seismic safety comm, 1983:81,86; Quarantelli, 1983:70).

EXAMPLE: Kansas City Hyatt Hotel Skywalk collapse. The City Health Department post-disaster review noted that coordination was never established over patients transported by a helicopter operated by one of the local hospitals. The crew reportedly failed to coordinate with those directing scene activities, including the ambulance dispatcher, the triage officer, or the site communications officer (KC Health Dept, 1981:7).

Effects of Search and Rescue Activities
The manner in which search and rescue activities are carried out has an important influence on triage. Search and rescue often becomes the initial contact point with the disaster victims. Therefore, those carrying out this activity generally influence how the disaster casualties enter the EMS system. When search and rescue operations are confused and uncoordinated, the flow of casualties into the EMS system tends to develop the same characteristics (Quarantelli, 1983:63,67).

Search and rescue is not always coordinated or carried out with significant input by those having emergency medical expertise (Quarantelli, 1983:66). In part, this is due to the large amount of search and rescue that is carried out by unofficial civilian volunteers, often family members and neighbors (Drabek, 1981:xviii,38,53,68,87,97,111,119). It is also due to the ambiguity regarding who has the overall responsibility to coordinate search and rescue operations (Quarantelli, 1983:67; Drabek, 1981:xx,35,240; Wenger, 1986:32).

Lack of Inter-organizational Planning
In many disasters, the flow of disaster casualties into the EMS system has not occurred according to any formal, predesignated plan. In the 1970's, the Disaster Research Center found casualty flow occurring according to plan in only about half the cases they studied. In part, this occurred because many communities did not have a realistic inter-organizational plan for disaster EMS. This was the case in over 66% of the localities studied. Furthermore, even when plans existed, they were often limited in scope, dealing with only a single jurisdiction, or calling for the coordination of only 2 or 3 of the community's emergency agencies. In only about 25% of the cases studied was there anything resembling a region-wide plan (Quarantelli, 1983:71,86,101,103,106). Although there have been improvements in disaster planning since the 1970's, difficulties continue to be seen (Wenger, 1986:ii).

Even when plans for triage exist, they may be "paper plans." That is, they are either unrehearsed, devoid of associated training, based on invalid assumptions, or encompass only a limited number of those who actually participate in the disaster response (Dynes, 1981:75; Quarantelli, 1985:7).

Lack of Needs Assessment
Efficient use of available medical resources (including hospital facilities) re-quires an overall needs assessment to determine the numbers, types, and severities of injuries. It is also crucial to ascertain the availability and status (available versus in-use) of medical resources such as field medical personnel (EMTs, paramedics, nurses, physicians); equipment; ambulances and rescue and/or first-aid vehicles; and hospital facilities. However, disaster response has often evolved without consideration of the overall situation. Rather, many individuals did what seemed rational from their own isolated perspective. That is, they endeavored to move each disaster victim to the closest hospital as quickly as possible. Even when triage did occur, most often it included only an assessment of specific individual casualties, rather than an evaluation of the. disaster as a whole Quarantelli, 1983:111; Golec, 1977:169).

When the medical aspects of situation analysis have been neglected, two factors in particular seem to have contributed to this oversight.

Lack of Medical Direction at the Scene
In some cases, those with emergency medical training have not played a major role in the overall direction of activities at the disaster scene. Others likely to be directing disaster site operations may lack familiarity with the function of emergency medical systems (EMS) Quarantelli, 1983:66; U.S. Fire Admin, 1980:3,18,21,28,41,43).

Lack of Scene-to-Hospital Communications
Communications between the scene and area hospitals is essential for situation analysis and casualty distribution, yet meaningful and informative scene-to-hospital information flow (see Chapter 5) is often neglected (Golec, 1977:174; Neff, 1977:186; Edelstein, 1982:159; Goodwin, 1982:14). The Disaster Research Center found less than 22% of the cases where meaningful information ex change occurred between the disaster site and any area hospital Quarantelli, 1983:67). Even the absence of elaborate, preplanned procedures for scene-to--hospital communications related to triage activities, was no guarantee that the procedures would actually be used.

EXAMPLE: in one community, the disaster plan included procedures to prevent the overloading of any single hospital. The central communications center had access to information on each hospital's bed census and emergency department capability. The communications center was to notify the hospitals in the event of a disaster and was to direct patients away from overloaded hospitals. In spite of the plan, 90% of the 140 casualties were taken to one hospital out of 17 in the community. The remaining 15 were distributed among three other hospitals. Furthermore, the communications center never even notified the hospital that the disaster had occurred (Golec, 1977:172).


IMPROVING TRIAGE

Many of the general principles applicable to disaster management in general may improve triage. Use of common terminology and the existence of joint planning, training, and testing all contribute to effective activity. Procedures for cooperative communications, situation assessment, resource management, and integration of unexpected or unfamiliar responders are all applicable to organized triage efforts.

Coordination with Non-Medical Organizations
Successful triage is dependent not only on the actions of medical (EMS) personnel at the site, but on non-medical responders as well. Often, the majority of casualties in disasters are initially encountered during search and rescue efforts. Although the very first search and rescue is usually carried out by civilians who happen to be in the impact area, when the activity is taken over by formal emergency responders, they are most likely to be firefighters or peace officers (Quarantelli, 1983:66). Triage is more successful if injured casualties located by search and rescue efforts are fed into the triage system. This requires a concerted effort by those overseeing these two essential activities.

Coordination is also important with other organizations whose activity might affect triage. Examples are those responsible for adequate crowd and traffic control, decontamination of those exposed to hazardous substances, and provision of light and shelter for triage areas.

Coordination with Hospitals
Notification of Hospitals
Functional procedures are required to designate a person whose responsibility it is: 1) to see that all area hospitals are notified that a disaster exists and provided with information regarding its location, character, magnitude, and the numbers, types, and severities of casualties to expect; 2) to continually and regularly update this information; 3) to respond to requests from the hospitals for further information; and 4) to indicate when the hospitals may deactivate their disaster status.

Hospital Capacity Assessment
In order to distribute casualties rationally among area hospitals, someone at the scene needs to be responsible for acquiring information from the hospitals regarding their capacities and capabilities. This information needs to be up-dated continually, because hospitals are also likely to be receiving casualties who have gotten there by their own means. In addition, as off-duty staff come in, the hospital may be able to care for more patients than when the facility was initially notified. In some communities, one of the local hospitals is designated as a "disaster coordination hospital," responsible for collecting capacity information from the hospitals and casualty information from the scene. This facility is then responsible for directing ambulance destinations based on this information.

Coordination of Scene Medical Activities
One model for the coordination of scene medical activities is that described in the 1986 version of the California Fire Chief's Association, Multi-Casualty Operational procedures (MCOP) Manual (CFCA, 1986). What follows is a brief description of the system. For more detailed information, a copy of the manual is available from the Association at 825 M Street, Rio Linda, CA 95673, $6.50. The system is designed as a medical component of the Incident Command System (ICS) (see Chapter 7), and uses procedures and terminology consistent with ICS. The organizational structure is diagrammed in Fig. 8-1. This structure may be expanded to encompass multiple triage areas as illustrated by the diagram in Fig. 8-2. Each position is provided with a checklist of responsibilities similar to that type of checklist used for the ICS. An example is given in Fig. 8-3. Similar checklists are provided for all positions in the Multi-Casualty Incident Procedures Manual.

(click to enlarge)
Figure 8-1. Organizational structure of triage. (Adapted from Multi-casualty incident operations procedures manual, Rio Linda, 1986, California. Fire Chiefs Association.)

(click to enlarge)
Figure 8-2. Expanded organizational structure for multiple triage sites.

MEDICAL GROUP SUPERVISOR
Definition:	Qualified officer.
Commanded by:	Division Supervisor, or Branch Director, or Operations Chief, or Incident Commander [whichever is the lowest position that is activated].
Subordinates:	Triage Unit Leader, Treatment Unit Leader, Medical Transportation Unit Leader, Medical Supply Manager, Morgue Manager.
Function:	Establish, command, and control the activities within a Medical Group in order to assure the best possible emergency medical care to patients during a multi-casualty incident.
Duties:	1. Establish and supervise a Medical Group at a level of personnel and other resources sufficient to handle the magnitude of the incident. 2. Delineate officers and designate patient control area locations as appropriate. Isolate minor treatment and morgue areas. 3. Ensure law enforcement/coroner involvement as necessary. 4. Ensure activation of hospital alert system. 5. Request Hospital Emergency Response Teams through the hospital alert system as necessary to provide medical assistance. 6. Determine amount and types of additional medical resources and supplies, e.g., Medical Strike Teams, Medical Task Forces, medical caches, ambulances, helicopters, and other methods of patient transportation. 7. Establish coordination of air ambulance (helicopter) operation between Medical Transportation Unit Leader and the Air Operations Director. 8. Establish liaisons with on-scene agencies, e.g., Coroner's Office, Red Cross, law enforcement, ambulance companies, county health agencies, etc. 9. Ensure that proper security, traffic control and access have been established. 10. Direct other medically trained personnel to appropriate unit leaders.

Figure 8-3. Medical group supervisor duty checklist. (Adapted from Multi-casualty incident operations procedures manual, Rio Linda, 1986, California Fire Chiefs Association.)

The flow of casualties and layout of triage, treatment, and transportation areas are illustrated in Figs. 8-4 and 8-5.


TRIAGE PROCEDURES

Examples of Triage Classification Systems
There is no single, standard, or universal method of triage. The number of categories used may vary from 2 to 5 or more, depending on the particular system in use. Various color codes, numbers, and symbols have been used to identify these categories. The triage category is often identified by the use of a triage tag, the design of which is also variable. In the absence of a triage tag, a triage symbol is sometimes written on the patient. The selection of how many categories or what colors or symbols to use for triage is somewhat arbitrary, and each system has its particular advantages and disadvantages. If the number of categories is limited to two, for example, the system is simple to remember. On the other hand, the use of more categories has the advantage of greater precision (Rund, 1981; Savage, 1977; Gazzaniga, 1979; Baker, 1979; Moore, 1967; Grant, 1982; Silverstein, 1984:12,63; Cohen, 1982a).

In order to illustrate how triage categories can be used, examples of two classification schemes will be given. The idea is not to endorse any particular system, but to illustrate a small sample of the various methods which demonstrate the basic concepts of triage.

(click to enlarge)
Figure 8-4. Multi-casualty scene. (Adapted from Multi-casualty incident operations manual, Rio Linda, 1986, California Fire Chiefs Association.)

(click to enlarge)
Figure 8-5. Patient Flow Chart. (Adapted from Multi-casualty incident operations manual, Rio Linda, 1986, California Fire Chiefs Association.)


The S.T.A.R.T. System
START is a mnemonic for Simple Triage And Rapid Treatment. This program was developed in southern California by a group of emergency physicians, firefighters, and an emergency nurse (Super, 1984).
The basic process for determining categories is represented by Fig. 8-6. Information and training materials for the START system may be obtained from:

S.T.A.R.T.
Hoag Memorial Hospital Presbyterian
301 Newport Blvd., Box Y
Newport Beach, CA 92663
(714) 760-5689

A Proposed 5-Category Triage System
An alternative proposed triage system is described below. This system was designed to address some problems associated with the triage and management of "unsalvageable" disaster casualties. It was also constructed to be adapted not only to disasters, but also daily EMS care. The daily use of routine triage for EMS helps to assure that it will be a familiar system when disaster strikes. This system uses five color-coded priority categories. These are felt to be a representative sample of what is in common use (Savage, 1977; Baker, 1979). However, other effective systems use three or four categories. A summary of the system is given in Table 8-5, and its special features are then discussed.

Used in this context, "simple care" is that which does not require unusual expenditures of time, equipment, or personnel. "Simple" field care might include inserting an airway, sealing a penetrating chest wound, applying a MAST unit, or giving intravenous fluids for shock. "Complicated" care would include artificial ventilation or CPR. "Simple" hospital care might include giving intravenous drugs, applying splints, surgical cleansing of flesh wounds, performing a cricothyrotomy or inserting a chest tube. In some cases, it might also include surgical exploration of the abdomen. It would not include repair of a transected aorta or ruptured aortic aneurysm, nor would it include surgery for a fractured neck. The "catastrophic" category deserves special discussion. In some triage systems this has been called the Expectant category. It was reserved for those who were moribund or who were in such poor condition that they could only be saved if extensive resources were diverted from more salvageable cases. Immediate care was to be withheld from these cases so that limited resources could be used to do the most good for the most casualties. These casualties were grouped with the dead and minor casualties and given last priority (Tintinalh, 1978). This assignment of last priority was probably a holdover from the days when much of the military casualty sorting was carried out at the battalion aid stations where resources were extremely limited (Rund, 1981). It had to be assumed under these conditions that these casualties would not survive. In the Vietnam War, however, helicopters evacuating casualties often bypassed the battalion aid stations and delivered the casualties directly to the hospital. At the hospital, all living casualties were initially considered potentially salvageable (US Dept Defense,1975; Rund,1981). This may be a more appropriate procedure for triage in the civilian setting.

(click to enlarge)
Figure 8-6. The START classification protocol. (Adapted from Super, G.: START instructor's manual, Newport Beach, Presbyterian Hoag Memorial Hospital.)

(click to enlarge)

Table 8-5. 5-Category Triage System

PRIORITY	COLOR	SYMBOL	CASUALTY CONDITION
FIRST	RED	R	CRITICAL: likely to survive if simple* care given within minutes.
SECOND	BLUE	B	CATASTROPHIC: Unlikely to survive and/or extensive or complicated care needed within minutes.
THIRD	YELLOW	Y	URGENT: Likely to survive if simple** care given within hours.
FOURTH	GREEN	G	MINOR: likely to survive even if care delayed hours to days. May be walking OR stretcher cases.
NONE	BLACK	X***	DEAD
*Simple: Care that doesn't require unusual equipment, or excessive use of time or personnel. **Assigned THIRD priority (after YELLOWS) when there are so many casualties that if resources are used in vain to try to save BLUE cases, the YELLOWS will needlessly die. ***The circling of this symbol prevents its being confused with a sloppily written Y.

Assigning last priority to Catastrophic cases and reserving that category for those with inevitably fatal conditions presents several problems:

• It is generally well adapted only for use in situations where there are truly massive numbers of casualties and access to extremely limited resources. These conditions are not at all typical of civilian disasters in the United States.
• Psychologically, it is untenable to "condemn" living casualties to such a category, especially under conditions of maximum duress, and by persons not experienced in making such decisions (Gann, 1979; Gazzaniga, 1979; Grant, 1982; Moore, 1967; Tintinalli, 1978). Fortunately, it is unlikely for any one person to encounter civilian mass-casualty situations frequently enough to become "experienced."
• Treating these patients last means that minor injuries (which may survive for days without treatment) are given priority over casualties that may still be salvageable once the resources needed to care for them are finally mustered.

Assigning last priority to catastrophic casualties, therefore, is probably not a realistic approach in the civilian setting. In addition, it does not work well if the triage system is to be used on a daily basis for routine emergency patients.

The approach described here, therefore, is to first treat the Critical (RED) casualties. After this, the Catastrophic (BLUE) casualties, which it may stiff be possible to salvage, are usually treated. When there are large numbers of casualties, it is conceivable that those initially categorized as Urgent (YELLOW) will eventually reach a point where they will need care within minutes. If it appears that this will occur before the Catastrophic (BLUE) casualties have all been treated, then attention is diverted from the BLUEs to first treat the YELLOWs. Minor (GREEN) casualties are not treated until attention has first been given to REDs, YELLOWs, and BLUEs.

Detailed examples of the patient problems in each category are given in Appendix D.

Triage Tags
It is commonly suggested that disaster casualties have their priority indicated by the attachment of a triage tag. There is no universal agreement regarding the design of such tags. Several useful variations are in use (Cohen, 1983; 1977; Cohen, 1986).

The S.T.A.R.T. System uses a commercial triage tag (METTAG) with four categories, indicated by four tear-off, colored strips at the bottom of the METTAG (see Table 8-6).


Table 8-6. The Triage Classification System Used by METTAG
GREEN: (Bottom strip)
Symbol: ambulance--crossed-out
Meaning: No hospital treatment needed; first aid only

YELLOW: (Second strip from bottom)
Symbol: Turtle
Meaning: Non-urgent; hospital care

RED: (Third strip from bottom)
Symbol: Rabbit
Meaning: Urgent; hospital care

BLACK: (Fourth strip from bottom)
Symbol: Cross/dagger
Meaning: Dead or unsalvageable; no CPR

(Adapted from: METTAG literature, Starke, FL)


There are several practical features of this tag. It is designed so that if casualty's condition deteriorates, the next strip can be torn off to indicate the fact. Each tag has an identification number on it and on each of the color strips. In addition, two upper corners of the tag have the number on it, and they can be removed by tearing along perforations, and used for keeping track of the casualties. The upper part of the tag has spaces for patient information. Two disadvantages have been noted with the METTAG: 1) some responders have complained that the colored strips on the METTAG are hard to see at a distance; and, 2) whereas a patient's deteriorating condition can be indicated on the METTAG by tearing off an additional strip, an improving condition cannot be so easily indicated.

The amount and type of casualty information to be placed on the triage tag is also by no means standardized. The tag may provide for notations regarding such things as locations and types of injuries (sometimes indicated with a diagram of the human body), pulse, respiration, blood pressure, treatment given, a serial number, and the patient's name, address, age, gender, and next of kin.

Another approach is used by the Alpine, Mother Lode, and San Joaquin Emergency Medical Services Agency in California. They use a tag for patient information, and a separate, colored cloth tag to indicate the triage category. Thus, when the casualty's condition improves or deteriorates, the cloth tag is changed and the information tag remains with the patient. These cloth tags are inexpensive, durable, and visible from a distance.

One problem that has been observed in disasters is that, while the disaster plan called for the use of triage tags, there was a lack of tags at the incident site (KC Health Dept, 1981:5; Buerk, 1982:643; Quarantelli, 1983:77; Worth, 1977:163). The best solution to this problem is to keep a set of triage tags aboard every emergency rescue and ambulance vehicle. Each batch of tags should also include a summary of triage plan and categorization scheme, as well as charts for keeping track of casualty conditions, hospital capabilities, and hospital destinations.


Casualty Distribution Procedures

Variations Depending on Local Conditions
There are a number of different approaches to disaster casualty distribution. Which approach is most practical may depend on the size of the community, the number of area hospitals, and the difference in capabilities of these institutions.

In the simplest case, only one local hospital is available. However, it may be necessary for this hospital to act as a triaging facility, stabilizing patients then distributing them to more distant facilities (Butman, 1982:140). If there are but a few hospitals in the community, all with similar capabilities, it might suffice to send one ambulance load to each facility on a rotating basis. In large urban centers, a rather sophisticated set of distribution procedures may be necessary.

Hospital Polling
In some communities, the disaster plan may include procedures for polling each hospital to obtain information about its present staffing, number of empty beds, operating room availability, and other resources. Such plans need to recognize the time it takes to collect this information and to consider the fact that casualties typically begin to arrive at the hospital within 30 minutes of disaster impact (Quarantelli, 1983:74; Golec, 1977:173). In order to be of maxi-mum use, the polling information has to be made available to field medical units before the casualties leave the scene.

The "First Wave" Protocol
In communities that use hospital polling, it may be advantageous to predetermine a method for equitably distributing the initial disaster casualties, pending the collection of hospital information. By its very nature, such a procedure is likely to be imprecise. Nonetheless, even a fairly crude distribution of casualties is better than what is often achieved if no procedure is in effect.

An example of such a procedure is the first-wave protocol. This involves the predetermination of disaster treatment capabilities of the area hospitals as a guideline for casualty distribution. These are based on the "worst-case" types and numbers of casualties each facility can treat (i.e., 2 a.m. on a Saturday). These categories are established to match those used in the local triage categorization system. Using the above five-tiered triage system as an illustration, the hospital capable of treating a minimum of three urgent ("yellow") casualties would be designated as a "yellow-3- first-wave facility. Another hospital, capable of treating only minor casualties, but which could manage 20 of them, would be designated a "green-20" first-wave facility and so on. Although priorities are different for "red" (critical) and "blue" (catastrophic) casualties, the facilities needed for treatment are the same. Therefore, red and blue casualties would be sent to facilities with a "red" designation. Communities may consider the use of guidelines established by the Committee on Trauma of the American College of Surgeons in the selection of facilities for critically injured casualties (ACS, 1986:4).

Using this protocol, a community distributes casualties according to the type and number in the first-wave designation of each facility. When, for example, a "red-Y' facility has received three critical casualties, any further critical victims are sent to other "red" facilities until all such hospitals have received their quota. A modification of this system (applied only to critical casualties) has been initiated in Sacramento County, California (Lowry~ 1983).

When all hospitals of any particular color designation have received their share of casualties, subsequent distribution is according to a calculated first-wave ratio. This is determined by adding up all the hospital capacities for a triage category and dividing by the number indicating the capacity of each. For example, if the community has a total of 10 "red" casualties, and Hospital A has a first-wave designation of 4, then 4/10 or 40% of the "red" casualties from the disaster are sent to this facility. First-wave designations for a hypothetical community are illustrated in Table 8-7. The application of this First-Wave protocol is illustrated in Fig. 8-7.

Unfortunately, the existence of a disaster does not necessarily diminish the occurrence of routine emergencies. People continue to have babies, get sick, get drunk, and crash their cars into each other. The first-wave protocol can be used to take into consideration all the accidents and illnesses occurring in the community, including those created by the disaster. In this case, the sum of all the critical casualties at the disaster site and occurring in other areas of the community are used to determine the total load of critical casualties to be distributed to "red" hospitals. In a similar manner, distribution is determined for other categories of patients both on and off the disaster scene.

As in the case of triage categories, distribution techniques can be adapted for use in routine emergencies. When any hospital in the community, as the result of one or several emergencies, receives simultaneously a total number of emergency patients exceeding its first-wave score, this can be used as a guide-line for considering temporarily redirecting ambulance traffic to other facilities. I Likewise, when all "red" facilities have patient loads exceeding their first-wave scores, then the first-wave ratios can be used as a guide to divvy up the patient load. The adaptation of such types of disaster procedures for more routine emergency situations, tends to assure that the users keep familiar with them.


SUMMARY

Triage, a wartime invention, involves the concept of "doing the most good for the most casualties." As such, it is well adapted for use in civilian disasters Triage is often thought of in narrow terms as merely the designation of priorities for patient care., However, in disasters, doing the most good for the most casualties also means maximizing the use of the available hospital facilities. This is often a difficult task to accomplish, especially in diffuse disasters covering a large geographic area. This chapter has examined some of the more common problems interfering with effective triage and casualty distribution in disasters. The reasons for these problems and some examples of how to counter them have been discussed.

Table 8-7. Calculating the First-Wave Score and Ratio

A	B	C	D	E
Designation	Hospital	First-Wave Score	Total First-Wave Score of All Area Hospitals*	First-Wave Ratio
Red	Hospital A	4	10	4/10=40%
Red	Hospital B	2	10	2/10=20%
Red	Hospital C	2	10	2/10=20%
Red	Hospital D	2	10	2/10=20%
Yellow	Hospital E	8	20	8/20=40%
Yellow	Hospital F	12	20	12/20=60%
Green	Hospital G	5	20	5/20=25%
Green	Hospital H	15	20	15/20=75%
*with the same "designation" Determine the highest triage category that the hospital can manage at any time of day or week, giving good disaster care for the initial 2 hours, without calling in disaster back-up resources. Place this designation in column A. Determine the maximum number of casualties that can be managed at one time - under such conditions. Place this number in column C. Determine the total number of casualties of this triage category that can be managed by all hospitals in the area. Place this number in column D. Divide the number in C by that in D (multiply the product by 100 to convert it a percentage). Place this ratio in column E.

(click to enlarge)
Figure 8-7. Distribution using the "First Wave" protocol


PLANNING CHECKPOINTS

• Do persons with expertise in emergency medical services have primary authority over patient care and transport at the scene of a disaster?
• Do your fire and police personnel understand that they may be involved in disaster search and rescue operations?
• Do they understand how they should coordinate their activities with those involved in triage at the site?
• Does the disaster plan and training provide for close contact between those directing search and rescue and those providing EMS?
• Does your community have a plan and associated training for disaster casualty distribution among area hospitals?
• Are your disaster triage and distribution procedures adapted so they can be used in more common emergencies?
• Does your community have procedures and training for assessing the types, numbers, and severities of casualties at the scene? For sharing this information with all the involved responding organizations including all area hospitals?
• Is this information collected and disseminated on an ongoing basis?
• Does your community have functioning procedures and training for assessing the capacity and capability of local hospitals?
• Does every ambulance and rescue vehicle have a supply of triage tags?
• Does your plan anticipate that "trauma centers" and hospitals near the disaster tend to get a disproportionate share of casualties?
• Are all public safety and EMS personnel required to have ongoing training familiarizing them with the local disaster triage and EMS system?
• Does your plan and training indicate who is responsible for each of the following field disaster medical care/triage responsibilities:
• Overall coordination?
• Liaison with other agencies? Dispatch?
• Provision of disaster information to hospitals?
• Assessment of up-to-date hospital capabilities? Casualty distribution?
• Triage?
• Patient care at the scene? Logistics?
• Air transport?
• Public information?

ADDITIONAL READING

Burkle FM Jr, Sanner PH, and Wolcott BW: Disaster medicine: application for the immediate management and triage of civilian and military disaster victims, 1984. Available from: Medical Examination Publishing Co, Inc, 3003 New Hyde Park Rd, New Hyde Park, NY.

Butman AM: Responding to the mass casualty incident: a guide for EMS personnel, 1982. Available from: Emergency Training, 181 Post Road West, Westport, Conn 06880.

Cohen E: A better mousetrap: what makes up the "perfect"triage tag?, J Emerg Med Serv, pp. 30-36, July 1983.

Cohen E: Patient identification: a look at triage tags, Emerg Med Serv, 15(9):45-49, Oct 1986.

Eisman B: Combat casualty management in Vietnam, J Trauma, 7(l):53-63, 1967.

Emergency War Surgery: First United States revision of the emergency war surgery NATO handbook,United States Department of Defense, 1975. Available from: Superintendent of Documents,US Government Printing Office, Washington,DC 20402.

Golec JA and Gurney PJ: The problem of needs assessment in the delivery of EMS, Mass Emergencies, 2:169-177, 1977.

Multi-Casualty Incident Operational Procedures Manual, 1986. Available from: California. Fire Chief's Association, 825 M Street, Rio Linda, Calif 65673.-.'

Quarantelli EL, Delivery of emergency medical services in disasters: assumptions and realities, 1983. Available from Irvington Publishers, 551 Fifth Avenue, New York, NY 10017.

Rund DA and Rausch TS: Triage, The C V Mosby Co, St Louis, 1981.

Super G, editor: S.T.A.R.T. instructor's manual, 1984. hospital Presbyterian, 301 Newport Blvd, Box Y,Available from: Hoag Memorial Hospital- Newport Beach, Calif 92663.

Trends in Triaging: A Random Survey of Triage Tags Used in the United States, Emerg I'rod News 88-91, Oct 1977.