Canadian computed tomography head rule and its impact on Singaporean practice

Abstract

Introduction: We aim to determine the usefulness of the Canadian Computed Tomography (CT) Head Rule in patients who present with minor and minimal head injury to the emergency department. Methods: Clinical information was retrospectively collected and the need for CT head was evaluated. The primary outcome measure was the requirement for neurological intervention. The secondary outcome measure was brain injury requiring admission or neurological follow-up. Results: A total of 1127 cases were reviewed. About 6.3% had clinically important brain injury; and 1.2% required neurological intervention. The high-risk factors were 45.2% sensitive (95% confidence interval [CI]=27.8-63.7%) and 76.2% specific (95% CI=73.5-78.7%) for predicting need for neurological intervention. All seven risk factors were 73.2% sensitive (95% CI=61.2-82.7%) and 69.8% specific (95% CI=66.9-72.5%) for predicting clinically important brain injury. Overall, the rule had a positive predictive value of 14.0 (95% CI=10.7-18.1) and negative predictive value of 97.5 (95% CI=96.0-98.4). Conclusions: In this retrospective validation of the Canadian CT Head Rule in the Singaporean context, we conclude that the lower sensitivity of the rule could be attributed to local neurosurgical practice, lack of application of the rule among clinicians and inclusion of patients with minimal head injury. Practically, the high NPV will mean that patients who do not fulfill the rule can be safely discharged with head injury advice, without the need for a scan. The judicious use of CT head can achieve savings for our health-care system.