“Triaging” Human Conditions
The word “triage” comes from the French word “trier”, meaning “to separate, sift, or select”. The concept of triage may have originated from the Napoleonic Wars. It was later adopted during World War I.
The use of an effective triage system during the time of war was critical and necessary. Assuming that all lives (bodies) are valued equally in the battlefield, how can the squad optimize the allocation of scarce resources (medical attention, medicine, beds/sanitary conditions, etc.) while under tremendous pressure? Those of you who practice Emergency Medicine might find this eerily familiar even today.
Basic battlefield triage systems hence dictated the following classifications:
- Those likely to survive, regardless of care received;
- Those unlikely to survive, regardless of care received;
- Those whose outcome can be significantly influenced by the care they receive. [This is where they’d direct most of the resources]
Rudimentary much? Indeed. In practice, this system is rarely followed. As you could probably tell, not all lives are treated equally in the battlefield (consider commanders vs. infantry units) and making a call on the “likelihood of survival” is…well, more of an art than science. Those in commanding positions could even prioritize the allocation of medical resources towards specific outcomes (preserving manpower vs. optimizing able-bodiedness).
Rise of the Emergency Room as “Catch-all” of Health Care Systems
Let’s fast-forward several decades. By the 1960s, health care delivery systems in the U.S. (and elsewhere) had evolved drastically. Patients are covered by 3rd party payors. Hospital-focused care delivery systems took prominence and patients with acute problems were directed to a special area called the “emergency room” (before there was the ER). As the demand for acute care increased drastically, the need to efficiently allocate hospital’s limited resources became increasingly critical. By the 1980s, emergency service had become one of the most important channels for patients accessing the health care system – a “catch-all” funnel, if you will. In fact, the U.S. Congress became so concerned about widespread reports of “patient dumping” that they passed the Emergency Medical Treatment and Active Labor Act (EMTALA), which has had a profound regulatory impact on all aspects of emergency care.
So where are we with Emergency Medicine today? Well, here’s a few facts:
Demand >>> Supply. Americans make 130 million visits to the Emergency Room annually, or roughly 42 visits for every 100 people we have here in the U.S. Take that in for a second. This has been a growing trend years in the making, and at the same time, the number of hospitals operating Emergency Departments has declined from more than 5000 in 1991 to fewer than 4000 in 2006. [Source] Why are EDs closing down if there’s such a need, you may ask? As it turns out, the ED business may not be entirely financially viable for many organizations. [Source]
Utilization by Specific Demographics. Among all those who utilize the ED, ED utilization rates are especially high among infants, people age 75 and over, nursing home residents, the homeless, African Americans, and individuals covered by Medicaid/SCHIP. [Source]
Complexity & Bias. Patients’ perceptions of the urgency of ED visits often differ from the judgment of clinicians. In a study by Gill, et al. conducted on patients in the waiting area of an urban teaching hospital, they found that 82% of patients classified by triage nurses as non-urgent believed that their condition was, in fact, urgent. Patients often come to the ED for a variety of complex and overlapping concerns that include the need to quickly relieve pain or discomfort and “making sure everything is OK.” Caretakers of young children express additional concerns. Some feel they need professional reassurance to deal with their child’s inability to express pain and other symptoms precisely. Others want to make sure they are not to blame for the child’s problem. Some describe their choice to seek immediate verification of non-urgency as a form of parental responsibility. As strange as it sounds, this problem creates an inherent demand for rapid confirmation and reassurance.
A Whole Lot of Waiting. Think about a time when you or someone close to you were admitted in the ED. How was it? Patients spend a lot of time waiting in the ED. On average:
- You would wait for 24 minutes before you ever see a provider;
- If you break your bone, well, tough luck, it’d take 54 minutes before you get your hands on any pain medication;
- If you finally got to see a provider and they decide that you should be admitted to the hospital, it’ll take 96 minutes before you actually get wheeled to the room;
- If you’re lucky and they decided that your condition isn’t serious enough to be hospitalized, you’d be hanging out in the ED for 135 minutes before you could go home.
There’s a host of findings in the Synthesis Project conducted by the Robert Wood Johnson Foundation, which I’d highly recommend for those of you interested in this problem space. Here’s also a separate report from the Heritage Foundation on a similar topic.
Evolution of Triage Systems
As Emergency Medicine evolved over time, so did medical triage systems. During that time, many medical triage systems were developed, tested, and adopted worldwide. In a review paper published back in 2010, Christ, et al. attempted to evaluate four of the most prominent triage systems in existence – Australian Triage Scale (ATS), Canadian Triage and Acuity Scale (CTAS), Manchester Triage System (MTS), and Emergency Severity Index (ESI) – based on parameters of “validity” and “reliability”. While MTS and ATS touted a “three level” triaging schema, CTAS and ESI emphasized the need for more granularity in a “five level” triaging schema. If we leave the debate of “effectiveness” aside, I’m sure we can all agree that the “sorting of human conditions” is a pretty challenging problem.
Let’s highlight the fundamental premise behind why triage systems are needed:
The challenge we have on hand, and for the near future, is to develop an effective triage system, which, as their primary function:
- rapidly identify those who require the best available response for the medical problem presented, and
- focus the response using a sensitive and specific system, in order to
- use limited resources most effectively, and to
- focus the delivery of those in need of care to “centers” where definitive care can be given in one move.
So far, we’ve established a few key assumptions:
- People look towards the ED as a primary way to get acute care they need;
- ED overcrowding is a real issue;
- The ED cannot, and should not, be the “catch-all” function for patients seeking care.
The real issue we have on hand isn’t “fixing the ED”. You can have all the ED in the world and still end up with a massively expensive, non-scalable, problem on your hand. It’s about directing individuals to the appropriate provider of care where definitive care can be given before these individuals self-triage themselves into the ED.
I’m not talking about the cases where the ED would be a clear destination (say, you found a pole that pierced through your lung). It’s about the cases where there isn’t a clear line to be drawn.
In an ideal world, there should be a centralized “command center” which directs individuals in a finite set of population to the right level of care. Patients should have their chief complaint fully addressed with a singular touchpoint, and in cases where follow-up is required, there should be a clear mission to address the remaining issues in the follow-up. This should all take place before anyone touches the care delivery layer (including the ED).
From the system design perspective, this may look like a massive optimization problem for the time between the presentation of chief complaint and resolution of the said complaint:
- Minimize system cost (resources) incurred;
- Minimize time to care delivery;
- Minimize time to resolution of chief complaint;
- Minimize touchpoints/referrals required;
- Maximize patients’ need for trust and reassurance;
- Maximize convenience to patients;
- Maximize perceived and measurable value to patients
And that’s why this problem is interesting for many of us.
How do you get this right?