A common theme among many of the literature resources reviewed for the development of this chapter is discussion regarding the lack of sufficient data on which to base evaluation of the efficacy of existing triage methods. Of the few articles that have attempted to validate existing triage methods, because there is little existing data available regarding outcomes from real-life utilization of triage methods to actual MCIs and disasters, most studies are based on data surrogates such as retrospective application of protocol assessment criteria to patients in trauma registries. Considering that prospective assessment of a particular triage method is likely impossible due to barriers in predicting disasters, lead time in training providers, and certain ethical challenges, the use of these surrogates for data and efficacy assessment are necessary.
In addition to the lack of adequate data, many articles also cited the lack of a universally accepted gold standard or outcome measure with which to compare various triage methods. However, several different concepts have been identified as critical variables that must be considered when evaluating or developing a triage method.
Frykberg discusses the concept of the critical mortality rate, the percentage of deaths only among the critically injured, suggesting that “the outcome of critically injured casualties is the best indication of the success of medical care in an MCI.” By using this measure, triage methods would be compared based on their ability to identify and correctly categorize the critically injured patients, and would be judged on this specific survival score rather than on the overall disaster mortality rate (which would include the on-scene/immediate deaths as part of the entire fatality census).5
Several factors may influence the critical mortality rate achieved by a particular triage method. Ideally a triage method would correctly categorize each patient 100% of the time. However, certain rates of undertriage, inappropriate assignment of critically injured victims with life-threatening problems to a delayed category, and overtriage, assignment of noncritical casualties to immediate care, often occur. Undertriage places critical patients at risk of not receiving appropriate priority for treatment and transport. This may occur when victims have somewhat innocuous appearing injury patterns externally, but have significant internal injury (ie, small penetrating trauma from shrapnel). Conversely, casualties that have severe external injuries but have a low likelihood of survival may be overtriaged to the immediate category, rather than an “expectant” category. In either case, overtriage will lead to the consumption of resources that would best be utilized to care for the true “immediate” patients. Of the two, studies of MCI bombing events indicate that overtriage has been shown to have a greater negative impact, illustrating “a direct linear relationship between the rate of overtriage and the critical mortality rate of survivors”.9
Both over- and undertriage can be an effect of the triage method or of the rescuer who is using the method. Intrarater reliability occurs when an instrument results in identical triage categorization if the same evaluator rates the same patient twice within a short time period.10 Interrater reliability occurs when an instrument results in identical triage categorization of the same patient when evaluated by two different raters.10 In a well-developed triage method, rates of intra- and interrater reliability would be high.
When authors discuss the “testing” of a triage method, it is important to understand exactly what is being tested. Is one testing whether the scheme can predict patient outcomes, whether providers use the scheme accurately, or whether use of the scheme improves outcomes? In other words, when looking at the patient outcomes when a particular triage method has been utilized, it may be difficult to separate whether there was a success or failure of the tool itself, or success or failure in the ability of the providers to accurately/correctly use the tool.
The concept of construct validity, the ability of a test or process to assess what it is intended to assess, can be applied both to the validity of the triage method and in the tools used to assess the effectiveness of the method.11 The construct validity of several primary triage methods (START, SMART, CareFlight) has been assessed in a few studies, but no such assessment has been applied to secondary triage methods.10
In the initial chaos of an MCI or disaster, a certain amount of inaccuracy of triage must be expected and accepted. This inaccuracy can be mitigated, however, by utilizing secondary and tertiary triage at each point in the patient evacuation process (ie, arrival at a treatment zone, just prior to departure from a treatment zone, upon arrival at a destination hospital, etc). Such serial reassessments can help increase the accuracy of their diagnosis, can increase triage accuracy, and decrease the rates of under- and overtriage.