What is Triage?
What is the fundamental purpose of triage? Is to do the “most for the most” by making the best use of available resources for the most people, or is it to identify and prioritize the sickest patient at your doorstep so that you can “save those with the most time-sensitive illness”. I want to challenge that the fundamental decision on the effectiveness of triage tool is not merely based on the reliability and validity of a tool; but a much deeper analysis of the philosophical and ethical debate regarding the context and the purpose of the triage process within that context.
Traditionally the fundamental purpose of triage by definition is to be able to do the “most for the most” in a context where demand exceeds the capacity to meet that demand. Even this definition is not sufficient: how we define “demand” “exceeds” and “capacity” differs vastly in different contexts.
Triage originated in this context during the eighteenth century where the chief surgeon of Napoleon’s Imperial Guard, Baron Dominique Jean Larrey, prioritized the wounded based on their need for surgical intervention rather than rank. The purpose of triage in such contexts is to rapidly identify the sickest in the crowd and ensure that order of treatment increases the likelihood of survival. This relationship is not always linear and may include some patients that are too sick to justify treatment priority as the likelihood of survival is poor.
Triage is the corner stone of practice in emergency care. However the concept of triage can be applied to numerous settings depending on the infrastructure of the local healthcare system and the need to prioritize limited resources. Given that sick and injured patients present at any time, and that providers are often faced with more patients than they have resources to deal with, some form of prioritization for care is required. It is a process whereby patients are prioritized into certain acuity categories, which dictates their place in the queue.
As a process, triage can take place in many different settings, not just in a dedicated triage area and should certainly not be limited to a geographical location. Triage is dynamic: a patient’s level of urgency of need for medical attention may vary with time. Periodic re-triage of patients who are waiting to be seen is therefore necessary to monitor acuity changes over time. The process involved varies contextually, depending on current demand and available capacity.
How good is a triage tool?
Numerous triage tools exist, and most triage tools are based on a list of clinical discriminators; some include individual vital signs, while others include early warning scores (EWSs) or symptom-based algorithms. Traditional evaluation of these tools has been quantified using measures such as reliability, validity and time metrics.
Reliability measures if the tool produces results in a reproducible fashion. Reliability is not an inherent property of a tool; rather it is an interaction between the tool, the group of people using it, and the context. In the absence of an objective criterion, reliability is an important performance indicator for evaluating measurement quality and reducing error. No matter which cadres of staff are using the tool, there needs to be limited inter- and intra-rater variability in the application of a particular tool by different raters in a particular context.
What are the issues? Well some studies definitely use the term sensitivity and specificity interchangeably with reliability. Reliability should really be just about the tool it self. Also a lot of studies compare the use of the tool in the end-users to physicians. I can safely say that I haven’t been to a single hospital where a physician is sitting at the front of ER performing triage and assigning categories.
Validity measures if a triage tool is actually measuring what it promises to measure independent of its reliability (i.e. is the triage tool correctly identifying the true urgency of patients). Important questions to consider when assessing the validity of a tool are: (i) how appropriate is this tool to the context? (ii) How appropriate is this tool to the culture? (e.g. in Islamic countries, the process of triage may take place very differently as females may only be seen, triaged and treated by female healthcare staff). However in recent years surrogate measures for validity have been utilized by researchers, for example (i) time to treatment for those determined to be critically unwell, (ii) disposition. The problem with the former surrogate marker is that it is impossible to blind study participants, and it is dependent on the reliability of the triage tool. Using disposition is a marker for severity of illness is extremely dangerous, as it is based on the subjective assessment of the healthcare provider, and highly influenced by external biases such as bed availability, local protocols, ability to pay for treatment and the wishes on the patient.
So may be it is time to:
- Look beyond traditional reliability and validity studies. When choosing a triage tool engage in a philosophical and ethical debate centered on the local context to choose the right tool for you.
- Lets do better quality studies assessing reliability in end users and find better markers to test true validity.