To CT, or not to CT, That is the Question.

A 13-month-old girl falls out of her high chair and strikes her forehead on the ceramic floor. She cries right away and vomits one time. A large area of swelling develops in the middle of her forehead. She is seen by a physician at an urgent care center who promptly refers her to the emergency department (ED) for further evaluation. A 12-year-old boy sustained a head-to-head collision during a soccer game yesterday. At the time of the injury, he lost consciousness for a few seconds. He sat out the rest of the game and has been complaining of headache, nausea, and difficulty concentrating since the injury. His primary physician refers him to the ED for an imaging study. 

Sound familiar? As emergency medicine (EM) providers, the scenarios in which we encounter children who have sustained blunt head trauma are infinite. Whether we are pediatric EM fellowship-trained or EM residency-trained providers, we have extensive experience with trauma, and we are experts in the evaluation and management of these children. Our foremost objective with these patients is to rapidly differentiate those children who have sustained a clinically-important traumatic brain injury (ciTBI) and those who have not. However, we have also been tasked with a responsibility to make a decision that has become increasingly more complicated: to CT, or not to CT.

Ahhh, the 64-slice CT scanner. This technologically elegant machine is both a blessing and curse. As a diagnostic test for identifying ciTBI, it encompasses the ideal characteristics of speed, accuracy, and precision. Of course, when something sounds too good to be true, it usually is. “Primum non nocere,” “First, do no harm.” There is increasing awareness that CT emits a focused, high dose of radiation to radiosensitive tissues, such as the brain. CT imaging of head-injured children has risks of radiation-induced malignancy. Another less recognized disadvantage of CT imaging with this population is that it is too good: it will not only identify ciTBI, but it will also find clinically-insignificant injuries. When this occurs, children are often subjected to hospital admission and possible reimaging, which of course will increase their cumulative radiation exposure.

In October of 2009, a landmark study was published in the esteemed journal Lancet to address this exact subject: Kupperman N., et al. Identification of children at very low risk of clinically-important brain injuries after head trauma: a prospective cohort study. Lancet. 2009 Oct 3;374(9696):1160-70.

This study sought to derive and validate prediction rules to identify children at very low risk for ciTBI after blunt head trauma for whom CT might be unnecessary. It was a multi-centered, prospective, cohort study that enrolled and analyzed data from more than 42,000 patients from 25 EDs of the Pediatric Emergency Care Applied Research Network (PECARN). This study is a work of statistical art. The data are clean, the power is sufficient, the confidence intervals are narrow, the results are impressive, and the conclusions are practice-changing. The analysis clearly creates decision rule algorithms for head-injured children <2-years-old and >2-years-old by developing prediction trees designed to maximize the sensitivity and negative predictive value. The PECARN study is the Holy Grail emergency providers have been waiting for to tackle the CT-or-not-CT dilemma in head-injured children. Again, however, when something sounds too good to be true, it usually is.

Based on the PECARN study data, CT could be avoided in approximately 25% of children <2-years-old and 20% of patients >2-years-old. The rates of imaging in the study centers were lower than national averages, indicating that these results may be even more profound in typical general emergency centers. I would be willing to bet that more than 18 months after the PECARN study’s publication, national CT imaging rates of head-injured children have not declined 20-25%. Why not? Why, when the data is so compelling, when a decision rule is designed so gracefully, why are these changes in practice not being implemented on an expansive scale?

The barriers EM providers face with this matter are many. In the current zero-miss tolerance medicolegal environment that we practice, the pressure to image with CT remains high. The pressure comes from many sources including parents, primary and urgent care physicians, and from the EM providers’ personal experience and comfort level. The discussion with parents regarding the risk of radiation vs. the risk of a ciTBI is well worth the time, and I would hope that most parents would not want to needlessly expose their children to radiation. Primary care physicians should not be expected to determine who requires imaging, and therefore referring these children to the ED for our evaluation is absolutely appropriate. However, the referrals of these children from primary physician offices and urgent care centers at times create an expectation that an imaging study must be done. After all, why else would they be sent to the ED? Finally, the individual EM provider, who has now assumed all of the risk included with this patient, must make the ultimate decision. This provider’s experience, comfort level with evaluating injured children, clinical judgment, and knowledge of the evidence will all help shape that decision.

CT imaging of head-injured children is an issue that crosses not only the clinical spectrum of emergency care, but also the much broader aspects of healthcare delivery during this revolutionary period. A dissertation can be written on the subjects of utilization, ED throughput and length of stay, cost and reimbursement, tort reform, and how they relate to CT imaging of head-injured children. Future questions that will be encountered include the medicolegal ramifications of radiation-induced malignancies. As EM providers, the PECARN study has equipped us with an invaluable instrument: solid evidence. It is our responsibility to incorporate that evidence into our practice and do what is right for our patients. “Primum non nocere,” “First, do no harm.”