Diagnostic Accuracy of the 2019 Traumatic Brain Injury Guideline for Risk Stratification in Predicting Abnormal CT Brain Findings in a Community Hospital Setting

Abstract

BACKGROUND: Traumatic brain injury (TBI) is a common presentation in emergency departments. Selective use of computed tomography (CT) of the brain is crucial, particularly in resource-limited community hospitals. The 2019 Thai TBI Clinical Practice Guideline recommends CT indications based on risk stratification. However, its real-world diagnostic performance remains uncertain.

OBJECTIVES: To evaluate the diagnostic accuracy of the 2019 Thai TBI Clinical Practice Guideline for risk stratification in predicting abnormal CT brain findings among TBI patients in a community hospital.

METHODS: A retrospective study was conducted at Phanom Sarakham Hospital, Thailand, from January to June 2025. Patients aged≥15 years with blunt head trauma who underwent CT brain were classified into five CPG-based categories. Radiologist interpretations served as the reference standard. Proportions of abnormal CT findings, likelihood ratios, and area under the receiver operating characteristic curve (AUROC) were calculated.

RESULTS: Among 276 patients, 95 (34.4%) had abnormal CT findings. Abnormal CT rates increased with severity: 0% (mild low-risk), 33.3% (mild moderate-risk), 28.8% (mild high-risk), 69.3% (moderate TBI), and 90.0% (severe TBI). Positive likelihood ratios ranged from 1.1 to 17.2, while the overall AUROC was 0.61.

CONCLUSIONS: The 2019 Thai TBI Clinical Practice Guideline for risk stratification demonstrates a clear severity-dependent increase in abnormal CT findings but only modest diagnostic accuracy. Low-risk patients were safely identified. However, CT yield among mild TBI—particularly moderate-risk—was relatively low. Further refinement and prospective validation in community hospital settings are warranted.

Thaiclinicaltrials.org number, TCTR20260321008