Accuracy of an International Classification of Diseases Code Surveillance System in the Identification of Traumatic Brain Injury

Abstract

Background: Use of International Classification of Disease (ICD) codes for traumatic brain injury (TBI) in healthcare administrative databases may underestimate the epidemiology of TBI. The present study examined the use of ICD-10 codes in a population-based New Zealand sample. Methods: TBI related ICD-codes within the New Zealand health database were examined for all incident cases from a population-based TBI sample (n = 1,369). Impact of age, gender, ethnicity, presence/absence of skull fracture, where the case was located (i.e., hospital) on receipt of an ICD code was examined. Results: Individuals who received a TBI-related ICD-code accounted for 18.6% of the studied population. The most frequent codes were brief loss of consciousness, head injury not otherwise specified, and concussion. Receipt of a code was not impacted by gender. Those with skull fracture (χ2 (1) = 69.983, p < 0.001), and/or attending hospital (χ2 (2) = 81.507, p < 0.001), and of older age (χ2 (1) = 56.473, p < 0.001) were more likely to receive a code. Conclusions: Reported incidence of TBI, when based upon health registration data, is likely to be a significant underestimate. Specific research needs to be conducted to identify the barriers to medical personnel implementing the ICD head injury codes and to identify ways in which the use of the codes can be improved.