The Nitty-Gritty of Clinical Decision Rules Guy Carmelli, MD*; Andrew Grock, MD; Enid Picart, BS; Jessica Mason, MD *Corresponding Author. E-mail: carmellig@gmail.com. 0196-0644/$-see front matter Copyright © 2018 by the American College of Emergency Physicians. https://doi.org/10.1016/j.annemergmed.2018.04.004 SEE RELATED ARTICLE, P. 703. [Ann Emerg Med. 2018;71:711-713.] Editor’s Note: Annals has partnered with EM:RAP, enabling our readers without subscriptions to EM:RAP to enjoy their commentary on Annals publications. This article did not undergo peer review and may not reflect the view and opinions of the editorial board of Annals of Emergency Medicine. There are no financial relationships or other consideration between Annals and EM:RAP, or its authors. ANNALS CASE How are clinical decision rules (CDRs) applied (and maybe misapplied) in emergency medicine? This question arose after we read the article by Babl et al. 1 Here, they compared clinician gestalt with 3 decision rules for head injury in children. Surprisingly, CDR use would not have increased sensitivity and may have led to increased use of imaging! Aren’t CDRs supposed to aid judgment and reduce unnecessary testing? We had hoped yes, but unfortunately, they sometimes lead us astray. CDRs, otherwise known as prediction rules or prediction models, combine multiple patient historical and examination variables, test results, and other disease characteristics to estimate the probability of either a diagnosis or a prognosis. 2 The term “rule” is a misnomer in that CDRs are not inflexible or absolute, but should function more to supplement clinical judgment. 3 Hence, many providers prefer the term “clinical decision tools.” (To avoid a superfluous acronym, we will stick with CDRs.) CDRs CAN BE A BIG HELP In many ways, medical decisionmaking boils down to playing the percentages and predicting risk of outcomes. Being able to do this for every disease process is difficult, and CDRs can factor in many different variables to reach a consistent outcome. 4 CDRs can also help by organizing which clinical features play a significant role in predicting bad outcomes, formalizing a standardized method for physicians to approach a disease process, serving as a great tool for new learners developing their clinical gestalt, and supporting clinical decisions in our documentation. Furthermore, several prediction models have been shown to be more accurate than clinical judgment alone. 4 NOTHING COMES WITHOUT RISK Although CDRs have many benefits, they also have some inherent problems. They may themselves have poor generalizability or weak external validity, or physicians may misapply them, confuse variables, or not know how to incorporate their own gestalt. 5 Let’s discuss some of the major pitfalls some people face with CDRs, and offer up some pearls as well! THE PEARLS AND PITFALLS Wrong Population CDRs are often made for specific patients in a specific population. If derived at a different hospital, with a different practice style, or with a different patient population, that CDR may not be right for you! For example, the pediatric blunt abdominal trauma CDR created by Holmes et al 6 excludes focused assessment sonography for trauma ultrasonographic results. Therefore, it may not apply to hospitals that regularly perform serial examinations on their patients for focused assessment sonography for trauma. Individual hospitals and practice environments should decide which CDRs are applicable to their setting. 7,8 Wrong Patient The derivation studies for CDRs have specific inclusion and exclusion criteria. 5 Keep in mind that the patient in front of you may have certain characteristics that would exclude him or her from the CDR: young age, malignancy, dementia, pregnancy, intoxication, anticoagulation, or immunocompromised state. For example, patients with penetrating trauma were excluded by both the CATCH rule (Canadian Assessment of Tomography for Childhood Head Injury) for head trauma and the National Emergency X-Radiography Utilization Study C-spine rule. 9,10 The specific inclusion criteria are important too because the CDR should not be generalized to patients who don’t Volume 71, no. 6 : June 2018 Annals of Emergency Medicine 711 EM:RAP COMMENTARY