Rethinking Triage: Could Intuition Outperform Formal Systems?
Very little has changed or improved in MCI triage in 40 years
🕖 Reading Time, 12 minutes
In 1986, an article was published discussing “new concepts in triage.”1 The authors identified three areas of concern:
- Difficulties performing triage at an MCI
- Challenges with triage tags for casualties
- Alternative techniques for triage
Nearly forty years later, the current literature still struggles with these same issues. Apparently, very little has changed or improved in MCI triage.
To paraphrase an article looking at triage in complex, coordinated terrorist attacks specifically, trying to overlay a triage system, which by definition is “ordered, inflexible and process driven” onto a “dynamic, complex, and uncontrolled environment,” is extremely challenging at best and guaranteed to fail at worst.2 These same problems seem to apply to more traditional MCIs as well.
If you want to see how START, SALT, and RAMP Triage compare in a mass casualty event, also in Gingerbread, see our prior article.
The issue with triage “systems.”
A structured survey answered by 464 EMS employees, employed by three different systems, noted a formal triage system was used in only 16% of actual MCI events.3
Even when rescuers use a triage system, evaluating each injured person takes 30-60 seconds to determine their triage category.4
In a retrospective review of 29 military MCIs, comprising over 1000 casualties, no formal triage algorithm was used. In 28 of the 29 events, casualties were sorted in a binary fashion: “dying now or dying later / urgent or non-urgent.”5
Very little data validates that triage systems actually work at MCI events. Prospective studies also indicate that “triage algorithms have limited accuracy and reliability in sorting patients.”6
So-called “over-triage,” where a casualty is placed in a higher acuity group, risks wasting time and resources on that casualty when they aren’t as sick as the others in that category. “Under-triage” also has issues, as the casualty was deemed not as sick as they actually are, leading to delays in care and likely worsens outcomes.7
In a study using six triage systems to evaluate over 500,000 trauma patients, none of the triage systems consistently predicted the casualty’s outcome best. The actual mortality rate of expectant / “black” category casualties varied between 59 to 90%. Having 41% of the expectant injured survive is a significant indictment that the triage system is failing to correctly identify the seriousness of a casualty’s condition.8
Despite many decades of discussion regarding which triage system is best, the continued existence and use of numerous systems indicates none have proven more usable or superior to the others. Perhaps this disconnect is the biggest indicator that formal triage systems just don’t work in actual MCIs.9
If a triage system is used, should ambulatory casualties without significant injuries be included?
In 1986, Vayer1 felt the answer was no. Even if the “walking wounded” were placed in a specific area after triage, they wouldn’t remain and are unlikely to wait to be treated while responders are dealing with sicker casualties. Rather, they will leave and find their own way to a hospital.
This problem occurred in the Route 94 shooting in Las Vegas, where huge numbers of casualties, some seriously wounded, did not wait at the scene to be triaged, but instead used ride-sharing apps for transport to medical facilities.
Shackelford10 agrees ambulatory casualties need to be removed from the scene rapidly to decrease chaos and potential delay they could introduce to rescuers needing to treat more seriously wounded casualties.
Triage tags
“The triage tag system represents a good theoretical construct that has, however, proved to be a failure in practice.”1 In 1986, Vayer questioned even if triage tags were used, did they add anything to the sorting process? They felt triage tags took time to fill out, which delays patient care and evacuation. Additionally, the tags often serve as de-facto definitive labels on the casualty, possibly discouraging responders from looking beyond the tag and re-evaluating the injured who are already “labeled.”1
A structured survey answered by 464 EMS employees (employed by three different systems) noted that triage tags were used in 92% of MCI drills but only 34% of actual MCIs.3
Retrospective studies show the same findings of poor triage tag usage in MCI casualty management. Other authors note limited benefits of triage tags secondary to loss, damaged, illegible writing, different tags between neighboring agencies, crew unfamiliarity, and obscuration by body fluid.3
Triage tags were placed on some casualties from the active violent incident in San Bernardino, California, in December 2015, where 36 were shot and 14 killed. The placement occurred after extrication to a treatment area. However, halfway through the process, a decision was made to stop because using the tags was “interfering with patient assessment and treatment.”2
Geographical triage
Regardless of how casualties are triaged, Vayer felt it was important to sort similar severity casualties together in the same physical location to facilitate resources being brought to them. This now seems to have consensus as a good practice.
Geographical triage has been utilized successfully in many events, most notably in the Sioux City airline crash in 1990, where despite city disaster planning calling for triage tag use, they weren’t secondary to concerns it would slow down casualty management. Ultimately, by geographically collecting the injured into groups based on acuity 59 critically injured passengers were evacuated in 40 minutes.
Unique triage systems, as opposed to “daily practice”
In a review of 51 older civilian MCIs, “in most cases, no triage occurred at all.” They observed when some kind of triage was conducted, it was performed essentially, that is, the same way EMS responders evaluated patients daily, and not with some system specific to an MCI.1
In a retrospective review of 29 military MCIs, comprising over 1000 casualties, no formal triage algorithm was used. In 28 of the 29 events, casualties were sorted in a binary fashion: “dying now or dying later or urgent versus non-urgent.”5
The Royal Victorian Hospital in Belfast managed 25 MCIs between 1969 and 1976, secondary to terrorist incidents. They found the closer casualties were managed to the daily EMS routine, the better. They felt trying to implement a triage system that was infrequently used and not consistent with “daily practice” consistently failed.1
Intuitive Triage
A study was conducted randomizing paramedics to START (the state-mandated triage system for the paramedics) or “intuitive” triage, where they were told to use their “intuition” and place casualties from a simulated active shooter event into traditional triage groups. The study was conducted with 47 to 51 casualties in each run. Although the study design has limitations, “intuitive” triage was, on average, 36 seconds faster per injured person than START. The intuitive triage agreed with START’s categorization in 80% of casualties. In the 20% when disagreement occurred, the paramedics intuitive triage was evenly split between over and under triagaging the casualties (50% of the time in each).7
This “intuitive” triage has been used during actual MCI events. In the San Bernardino, California active violent incident in December 2015, where 36 were shot and 14 killed, although responders were trained in a specific triage system, they did not use it. Instead, they relied on “clinical judgment.”2
Shackelford proposes separating casualties into three groups: unstable, stable, and ambulatory (U-S-A triage).10 These categories are already familiar to EMS providers and essentially used in daily practice. If you remove the ambulatory category, injured rapidly divide themselves into sick and not-sick groups. Others have used “transport now, transport later, and dead.”6
The unstable or transport-now casualties can best be described as needing emergency surgery and blood transfusion to survive.5
With an overwhelming number of casualties, sorting strictly by mental status, “alert or not” may be the most operationally useful technique.5 Use of GCS alone has shown a strong predictive value for mortality in a review of over 500,000 trauma patients.8
Expectant casualties
In none of 29 military MCI events were any casualties declared expectant, although the obviously dead were identified as such.9 US military medics have said in combat zones they will not use an expectant category, even at an MCI.5 This approach will become particularly problematic in an ultra-MCI with hundreds to thousands of casualties. Subsequently, discussions have been generated about how best to teach military medics to declare appropriate casualties as expectant and beyond salvage with the resources available to the providers.11
Ultra-MCI - Is triage even feasible?
It is generally agreed upon that triage systems based on some vital sign can easily take 30-60 seconds for each injured person. While this may be feasible with a limited number of casualties, it rapidly becomes impossible with hundreds of wounded.
As the US military once again contemplates peer-to-peer or near-peer conflict, the concept of an ultra-MCI with hundreds to thousands of casualties will render any planned triage system moot. As Shackelford acknowledged, time will be the triage tool. Those injured but surviving long enough for rescuers to get to them, by definition, won’t have been the sickest, as those casualties will already die before care arrives.10
Principles of time as a triage tool shows that external hemorrhage control and airway clearance need to be completed within minutes of injury, blood transfusion within about 30 minutes, and the delivery of the wounded to surgical resources within an hour if the casualty needing those interventions is to survive.10
In an ultra-MCI, the first few minutes of intervention will have to be performed by bystanders and those present at the scene when the incident occurs. With hundreds or a thousand casualties, responders will not be at the side of rapidly bleeding or airway obstructed injured persons soon enough to intervene.
Review the Co-TECC Active Bystander Guidelines for more information.
References
1Vayer JS, Ten Eyck RP, Cowan ML. New concepts in triage. Ann Emerg Med. 1986 Aug;15(8):927-30. doi: 10.1016/s0196-0644(86)80677-1. PMID: 3740579.
2Pepper M, Archer F, Moloney J. Triage in Complex, Coordinated Terrorist Attacks. Prehosp Disaster Med. 2019 Aug;34(4):442-448. doi: 10.1017/S1049023X1900459X. PMID: 31389325.
3Ryan K, George D, Liu J, Mitchell P, Nelson K, Kue R. The Use of Field Triage in Disaster and Mass Casualty Incidents: A Survey of Current Practices by EMS Personnel. Prehosp Emerg Care. 2018 Jul-Aug;22(4):520-526. doi: 10.1080/10903127.2017.1419323. Epub 2018 Feb 9. PMID: 29425472.
4Cone DC, MacMillan DS. Mass-casualty triage systems: a hint of science. Acad Emerg Med. 2005 Aug;12(8):739-41. doi: 10.1197/j.aem.2005.04.001. PMID: 16079427.
5Rush SC, Lauria MJ, DeSoucy ES, Koch EJ, Kamler JJ, Remley MA, Alway N, Brodie F, Foudrait A, Barendregt P, Atkins M, Miller K, Hines R, Champagne M, Paladino L, Shackelford SA, Miles EA, Obiajulu J, Dorlac WC, Gurney JM, Robb D, Kue RC. Limitations of Triage in Military Mass Casualty Response: A Case Series. J Spec Oper Med. 2024 Oct 2;24(3):62-66. doi: 10.55460/0GO5-QW03. PMID: 39172917.
6Kamler JJ, Taube S, Koch EJ, Lauria MJ, Kue RC, Rush SC. Effectiveness of and Adherence to Triage Algorithms During Prehospital Response to Mass Casualty Incidents. J Spec Oper Med. 2023 Mar 15;23(1):59-66. doi: 10.55460/73Y0-FSLB. PMID: 36853853.
7Hart A, Nammour E, Mangolds V, Broach J. Intuitive versus Algorithmic Triage. Prehosp Disaster Med. 2018 Aug;33(4):355-361. doi: 10.1017/S1049023X18000626. PMID: 30129913.
8Cross KP, Cicero MX. Head-to-head comparison of disaster triage methods in pediatric, adult, and geriatric patients. Ann Emerg Med. 2013 Jun;61(6):668-676.e7. doi: 10.1016/j.annemergmed.2012.12.023. Epub 2013 Mar 7. PMID: 23465555.
9Rush SC, Lauria MJ, DeSoucy ES, Koch EJ, Kamler JJ, Remley MA, Alway N, Brodie F, Barendregt P, Miller K, Hines R, Champagne M, Paladino L, Shackelford SA, Miles EA, Dorlac WC, Gurney JM, Robb D, Kue RC. Rethinking Prehospital Response to Mass Casualty Events: Move, Treat, and Transport. J Spec Oper Med. 2024 Oct 2;24(3):24-29. doi: 10.55460/X38F-P3RH. PMID: 39271298.
10Shackelford SA, Remley MA, Keenan S, Kotwal RS, Baker JB, Gurney J, Rush S, Friedrichs P. Evidence-based principles of time, triage and treatment: Refining the initial medical response to massive casualty incidents. J Trauma Acute Care Surg. 2022 Aug 1;93(2S Suppl 1):S160-S164. doi: 10.1097/TA.0000000000003699. Epub 2022 May 23. PMID: 35583968.
11Cole R, Keenan S, Tadlock MD, Grover S, Givens M, Rudinsky SL. Expectant Casualty Care Training Needs for Future Conflicts. Mil Med. 2024 Aug 13:usae389. doi: 10.1093/milmed/usae389. Epub ahead of print. PMID: 39136473.
