Our scoping review found 55 studies on prehospital triage tools published within the past decade. These tools focused on general undifferentiated, trauma, and stroke populations and all included studies originated from high-income countries. Studies predominantly sought to assess predictive accuracy of the triage tools compared to in-hospital clinical outcomes, and many studied accuracy in simplified versions of existing tools. These published triage tools are generally designed to help prehospital providers determine destination of transport, means of transport and level of acuity. These tools also appear to provide a shared language for prehospital personnel to communicate with other emergency personnel, and assist in identifying vital sign derangements and exam findings across a spectrum of age ranges to differentiate ‘acute’ and ‘non-acute’ patients.
Trauma and stroke tools comprised over two-thirds of the included articles, perhaps because of their clinical and health systems significance. Outcomes for trauma and stroke depend on timely field recognition and are influenced by highly time sensitive interventions that are destination-dependent [45, 46]. Further, trauma and stroke care are regionalized in many high-income countries, therefore right patient destination are important to study for trauma and stroke system optimization. Last, both stroke and trauma outcomes are used to drive ‘benchmarking’ for health system accreditation and funding, which may also drive their importance as a research topic [47, 48, 49].
In trauma, the US FTDS appears to be the “industry standard” triage tool used, likely reflecting that the majority of our studies were from North America, specifically, the USA. As the majority of tools within the trauma triage literature derive from the FTDS, this well-researched tool is a promising starting point for further simplified trauma triage tool development, such as identifying individual components that may predict clinically relevant trauma outcomes. The trauma literature was relatively cohesive in that most studies used common clinical end points, which facilitates comparisons across studies.
In stroke care, while no single tool emerged as the prehospital triage ‘gold’ standard, the RACE, FAST-ED and Cincinnati Prehospital Stroke scales appear to have the highest quality data supporting their use. The National Institutes of Health Stroke Scale was presented in multiple studies as the gold standard in-hospital tool which was used for comparison.
The all-comer triage literature includes a myriad of tools with varying complexity, from those that incorporate vital signs alone (e.g., NEWS), to those with complex diagnostic algorithms incorporating history and exam findings to arrive at a level of acuity designation (e.g., CTAS). No one tool emerged as a clear gold standard, and authors’ use of a wide variety of clinical end points which make cross comparisons challenging.
Research themes common to these studies include simplifying existing tools such that they are efficient and accurate for the EMS provider to derive an accurate triage decision, and to identify the most accurate tool out of a large cadre of tools currently available. Standardized reporting of clinical end points would facilitate this endeavor in future research. Additionally, we noted a paucity of articles researching implementation or assessing end user perspectives [28, 50, 51], and no studies examined costs associated with triage decisions. Qualitative studies, cost analyses, and implementation studies would be helpful to further our understanding of the value provided by prehospital triage tools.
Lastly, all the studies included in this scoping review were performed in a few high-income settings, and the tools may not translate well to other high-income settings or LMICs with a different healthcare configuration, infrastructure and cadres of prehospital providers. Destination decision making would need to be locally-determined, especially in LMICs where specialty diagnostic (e.g., computed tomography scanners) and therapeutic resources (e.g., tPA) may be even more scarce. Further, triage tools may need to be tailored based upon regional injury and illness patterns. For example, prehospital triage of obstetric emergencies was notably missing from our review. Jenson et al. performed a systematic review of emergency department (i.e. in hospital) triage tools in LMICs and identified the South Africa Triage Scale (SATS), modified Early Warning Score and the Australasian Triage Scale as promising tools that had been validated across multiple studies in LMIC settings [8]. SATS has been implemented in the prehospital setting in South Africa and studies analyzing performance characteristics, while on-going, are yet to be published [52].
In recent years, prehospital care has received increased recognition in international health policy. Data extrapolated from the Global Burden of Disease study show that 24 million lives are lost each year in LMICs due to conditions sensitive to prehospital and emergency care. Ischemic heart disease, cerebrovascular accidents, and unintentional injuries are the largest contributors to morbidity and mortality in these settings [6, 53]. In 2019, delegates to the 72nd World Health Assembly adopted a resolution to strengthen emergency and trauma care systems and prehospital care was highlighted as an essential component [54]. Prehospital triage tools are a key building block for quality and safety assurance in the development of novel EMS systems [55]. It is our hope that this scoping review has provided a valuable framework for what is known thus far, and that further research will be done to advance the field.
The authors acknowledge the following limitations of this scoping review. First, the review was limited to English language publications. This may have excluded triage tools published in non-English journals. Secondly, the review was limited to only peer-reviewed published literature; it is likely that white papers and other non-peer-reviewed papers discuss additional triage tools currently in use. Lastly, inherent to this study’s design as a scoping review, the authors were unable to draw quantitative conclusions about the performance characteristics of the tools presented.