Elsevier

Annals of Emergency Medicine

Volume 64, Issue 2, August 2014, Pages 145-152.e5
Annals of Emergency Medicine

Pediatrics/original research
Comparison of PECARN, CATCH, and CHALICE Rules for Children With Minor Head Injury: A Prospective Cohort Study

https://doi.org/10.1016/j.annemergmed.2014.01.030Get rights and content

Study objective

We evaluate the diagnostic accuracy of clinical decision rules and physician judgment for identifying clinically important traumatic brain injuries in children with minor head injuries presenting to the emergency department.

Methods

We prospectively enrolled children younger than 18 years and with minor head injury (Glasgow Coma Scale score 13 to 15), presenting within 24 hours of their injuries. We assessed the ability of 3 clinical decision rules (Canadian Assessment of Tomography for Childhood Head Injury [CATCH], Children's Head Injury Algorithm for the Prediction of Important Clinical Events [CHALICE], and Pediatric Emergency Care Applied Research Network [PECARN]) and 2 measures of physician judgment (estimated of <1% risk of traumatic brain injury and actual computed tomography ordering practice) to predict clinically important traumatic brain injury, as defined by death from traumatic brain injury, need for neurosurgery, intubation greater than 24 hours for traumatic brain injury, or hospital admission greater than 2 nights for traumatic brain injury.

Results

Among the 1,009 children, 21 (2%; 95% confidence interval [CI] 1% to 3%) had clinically important traumatic brain injuries. Only physician practice and PECARN identified all clinically important traumatic brain injuries, with ranked sensitivities as follows: physician practice and PECARN each 100% (95% CI 84% to 100%), physician estimates 95% (95% CI 76% to 100%), CATCH 91% (95% CI 70% to 99%), and CHALICE 84% (95% CI 60% to 97%). Ranked specificities were as follows: CHALICE 85% (95% CI 82% to 87%), physician estimates 68% (95% CI 65% to 71%), PECARN 62% (95% CI 59% to 66%), physician practice 50% (95% CI 47% to 53%), and CATCH 44% (95% CI 41% to 47%).

Conclusion

Of the 5 modalities studied, only physician practice and PECARN identified all clinically important traumatic brain injuries, with PECARN being slightly more specific. CHALICE was incompletely sensitive but the most specific of all rules. CATCH was incompletely sensitive and had the poorest specificity of all modalities.

Introduction

Computed tomography (CT) is the criterion standard for diagnosing traumatic brain injury. Although it rapidly and accurately identifies traumatic brain injuries, potentially reducing morbidity and mortality, it is costly, may be difficult to obtain for children, and exposes patients to radiation.1, 2

Editor's Capsule Summary

What is already known on this topic

Some advocate clinical decision rules to guide head computed tomography (CT) use in injured children.

What question this study addressed

How do 3 clinical decision rules compare with clinical judgment?

What this study adds to our knowledge

In this prospective study of 1,009 injured children, only physician baseline ordering practice and Pediatric Emergency Care Applied Research Network (PECARN) identified all of the 21 clinically important brain injuries, with PECARN being slightly more specific. Physician risk estimation missed 1 injury, and 2 other decision rules were insufficiently sensitive.

How this is relevant to clinical practice

Baseline physician ordering practice and PECARN outperformed 2 other clinical decision rules for head CT use in injured children.

Clinically important intracranial injuries are rare, occurring in less than 5% of children presenting to the emergency department (ED) with minor head injury (Glasgow Coma Scale [GCS] scores of 13 to 15), and injuries requiring neurosurgical intervention occur in less than 1% of children.3 Decision analyses suggest that for most children who are at low risk of traumatic brain injury, the risks of radiation outweigh the risks of traumatic brain injury, and CT is not warranted.4 Despite this, more than one third of children with minor head injury undergo CT.5

Clinical decision rules may prove useful for guiding clinical decisionmaking for children with minor traumatic brain injury. Three recently published rules, the Pediatric Emergency Care Applied Research Network (PECARN) rule, Canadian Assessment of Tomography for Childhood Head Injury (CATCH), and the Children's Head Injury Algorithm for the Prediction of Important Clinical Events (CHALICE), show promise for improving clinical decisionmaking after minor head injury by potentially increasing recognition of injuries and reducing the frequency of CT acquisition (Table 1).3, 6, 7 For each of the rules, the absence of any features of the rule obviates the need for CT by categorizing a patient as low risk for clinically important traumatic brain injury.

Before being incorporated into usual practice, clinical decision rules require external validation and comparison to clinical judgment.8 Without validation, results may represent unique aspects of the studied patient population, clinicians using the rule, or overfitting of the model. Several examples exist in the literature of rules that worked well in derivation but failed when applied to new cohorts.9, 10, 11 Appropriate validation requires both an assessment of rule performance in settings apart from the derivation study and a comparison of performance to physician estimates of injury.12 To date, external validation and comparison have not yet occurred for the minor head injury rules.13, 14

In this study, we aimed to evaluate the diagnostic accuracy of PECARN, CATCH, CHALICE, and physician estimates for identifying clinically important traumatic brain injuries in children with minor head injury.

Section snippets

Study Design and Setting

We performed a prospective cohort study of children younger than 18 years and presenting to the ED at Denver Health Medical Center with minor head injury from January 15, 2012, through June 15, 2013. Denver Health Medical Center is a 477-bed urban, Level II pediatric trauma center for Denver, CO. The ED has approximately 30,000 annual pediatric visits managed by general pediatric, pediatric emergency, and emergency physicians. The ED did not participate in the original derivation of PECARN,

Results

During the study period, 1,526 children with head injury presented to the ED, and physicians completed data forms for 1,062 (70%) patients (Figure 1). Characteristics were similar between enrolled and nonenrolled patients: respectively, median age 6.1 and 5.0 years; sex 64% and 58% male; and GCS score 15 for 95% and 99%. We excluded 53 enrolled patients; nearly all were excluded because they presented greater than 24 hours after the injury. Nearly half of patients were discharged after initial

Limitations

The prevalence of clinically important traumatic brain injury is inherently low in children, and we identified only 21 instances of it. As a result, our estimates are less precise than those of the original derivation studies, preventing absolute conclusions about the exact sensitivities of the rules.

We also did not enroll all patients, potentially leading to selection bias. We did not identify any demographic differences between enrolled and nonenrolled patients. In review of the medical

Discussion

To our knowledge, this study, conducted in an urban medical center with a designated pediatric ED, is the first study to prospectively evaluate PECARN, CATCH, and CHALICE head-to-head. Our study is also the first, to our knowledge, to compare the 3 rules against physician estimation to understand how judgment compares with empirically developed tools. In our cohort, the PECARN rule and physician practice were the only approaches to demonstrate 100% sensitivity for identifying patients with

References (18)

There are more references available in the full text version of this article.

Cited by (0)

Supervising editor: Steven M. Green, MD

Author contributions: JSE, KB, JD, and MM were responsible for conceptualization of the study. JSE and JSH were responsible for study design. All authors were responsible for data collection and approval of the final article. JSE, EC, and JSH were responsible for data analysis. JSE was responsible for drafting the article. KB, JD, MM, and EC were responsible for article revision. EC and JSH were responsible for critical revision of the article. All authors had full access to all of the data (including statistical reports and tables) in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. JSE takes responsibility for the paper as a whole.

Funding and support: By Annals policy, all authors are required to disclose any and all commercial, financial, and other relationships in any way related to the subject of this article as per ICMJE conflict of interest guidelines (see www.icmje.org). Dr. Easter (K12HS019464) and Dr. Haukoos (K02HS017526) had financial support from the Agency for Healthcare Research and Quality for the submitted work. Dr. Easter was also supported by National Institutes of Health (NIH)/NCATS Colorado Clinical and Translational Science Institute grant UL1 TR001082. No authors had financial relationships with any organizations that might have an interest in the submitted work in the previous 3 years.

Contents are the authors' sole responsibility and do not necessarily represent official NIH views.

Please see page 146 for the Editor's Capsule Summary of this article.

A feedback survey is available with each research article published on the Web at www.annemergmed.com.

A podcast for this article is available at www.annemergmed.com.

View full text