Taking RiskThis post was previously published on iTeachEM on 9/22/14 here.

A 44-year-old healthy man presents with dull chest pain for 3 hours. His EKG is unremarkable. What’s his risk for acute coronary syndrome? Should he get a troponin? Two troponins? Observation and a stress test?

Emergency Medicine is an inherently risky specialty. In fact, many would say that risk stratification is our specialty. When a patient presents with symptoms, we use our clinical knowledge to determine what we think to be the most likely cause of those symptoms. We then apply studies and investigations to help confirm that diagnosis while attempting to “rule out” other diagnoses. At the end of this, we are often left without a specific diagnosis and need to make a disposition. When we decide to admit or discharge a patient, have them follow up in 24 hours or 1 week we are risk stratifying. For those we send home without a diagnosis, we try to determine how long they can wait to see another doctor for further investigation. We know that some of these patients will decompensate and return to the ED and so we are risk stratifying the likelihood of that decompensation.

Thus, during each patient encounter, the Emergency Physician needs to perform multiple risk stratifications. For example:

A 41-year-old man on aspirin presents with minor head trauma. His GCS is 15 and he is neurologically intact. He complains of a mild headache.

  • Does the patient need imaging now?
  • Does the patient need observation for 2 hours? 4 hours? Overnight?
  • Can I send the patient home safely without imaging?
  • Will the patient’s status degrade in the next 24 hours? 48 hours?
  • Should I schedule neurology follow up? If so, when?

This is a fairly simple case yet multiple risk assessments are involved. Each of these decisions must take into account hospital factors (i.e. ability to obtain follow up) and patient factors (i.e. distance from the hospital, reliable to follow up).

This brings us to the central questions of this post:

  • How do I train residents about risk?
  • How do I train residents to develop their risk threshold?
  • How do I train residents to embrace risk?
Damn it Jim, I'm a doctor not a CT scanner!

Damn it Jim, I’m a doctor not a CT scanner!

Clearly, we can see the need for this type of training. While we’d all like to have the magic tricorder to tell us if the patient has an intracranial injury, has a concussion etc we don’t. We deal with tests that are less than perfect and make decisions based on these tests.

This raises the first point I always discuss with my residents. There is no such thing as a “rule-out” test. There is no test or series of tests that can definitely “rule-out” a disease. We use the tests to risk stratify the patient. Take another case:

A 22-year-old woman presents with right lower quadrant pain and vomiting. She is tender and you order a CT scan of the abdomen and pelvis. The scan is read as “no intra-abdominal pathology is identified that explains the patients pain.” Has the patient been “ruled-out” for appendicitis?

We know the answer to this question is no. CT scan of the abdomen and pelvis has a sensitivity of 98-99% and so there will be patients that are false negatives. In spite of the fact that we know this, we usually tell patients, “You don’t have an appendicitis. You’re going to be fine and we’ll be discharging you in a bit.” What we should be telling patients is “The CT scan doesn’t show signs of appendicitis. I think it’s unlikely you have an appendicitis but the test isn’t perfect. There’s still a chance. We’re going to send you home but here’s what you need to watch out for.” The second statement is an acknowledgement that we have risk stratified to a low risk category but not no risk. This approach goes for any patient we see whether it be chest pain, an ankle injury or abdominal pain. Although this may appear to be nothing more than semantics, I argue that this change in terminology is central to teaching what Emergency Medicine is about.

Once we’ve rid the residents of the idea of ruling out disease, we need to encourage them to think about risk stratification when they present the patient and incorporate that into their presentation. Residents are smart and once they get to know the faculty, they tailor the presentation and their proposed workup to what they think the faculty member will want to hear. We should encourage them, instead, to present the patient and tell us what they would do if they were in charge. This allows them to begin to feel the responsibility of their plans. Unfortunately, this takes time. After they give you their plan, you need to explain why you would do things differently. Why is your plan more or less risky than that of the residents? Explaining this will allow them to develop their risk taking behavior.

The most important part of risk stratification and risk decisions is the patient. When I finish a patient encounter and am ready to discharge patients home, I always sit down and have a discussion about risks and the need for follow-up. These two things go hand in hand. Often, patients believe that discharge from the Emergency Department comes with a clean bill of health and a 5 year, 100,000 mile guarantee. This again reflects the disconnection between what we as physicians think and what patients perceive.

How do we teach residents to communicate risk to patients? First we should start with modeling the behavior. Have the resident follow you while you have this discussion with the patient. Here are some simple things to do to maximize this interaction and model the proper behavior:

  • Sit down and turn off the pager/phone (no interruptions)
  • Explain everything that’s happened during the ED stay
  • Explain the findings (or lack there of) from your evaluation
  • Discuss your evaluation of all of the information and the presence of clinical uncertainty and the importance of prompt follow up
  • Discuss how the follow up will be arranged (patient calls or you are calling for the appointment)
  • Ask if the patient has any questions

I also like to add, “I’m okay with being wrong but I want you to give me the opportunity to make it right. Come back and see me or one of my colleagues if anything concerns you.”

There is no act of teaching that benefits the resident more than watching the proper behavior modeled. For the next patient, you flip the scenario and have the resident lead the discussion and you watch them. Afterwards, you offer them critique and tips to improve.

Finally, I think there’s an important role in discussing difficult cases where your risk stratification was incorrect. I’m not talking about during formal department Morbidity and Mortality conference but rather conversations in the clinical environment about these cases. This act stresses the importance of following patients up in order to evaluate the appropriateness of your level of risk taking behavior. Residents should understand that our understanding and practice of medicine is not perfect and mistakes will be made. The vital thing is to learn from these mistakes and adapt our clinical care and risk taking behavior accordingly.

Risk stratification and risk taking behavior are central aspects of Emergency Medicine. It is our job as resident educators to help residents develop these skills and attitudes. Since I’m by no means an expert in this area, I encourage you to email me, post comments etc on the topic so we can all learn more.