Preliminary Evaluation of the Stockholm Score for CT Lesion Detection in Mild TBI

Overview

This preliminary analysis of 5,000 mild traumatic brain injury (mTBI) patients evaluates the Stockholm Score's ability to predict intracranial lesions (ICLs) on CT scans. The study aims to develop individualized risk models for traumatic ICLs and clinically significant lesions requiring intervention in the emergency department setting.

Background

Traumatic brain injury is a leading cause of morbidity and mortality globally, with mild TBI (GCS 13–15) comprising 70–90% of cases. While only a small fraction of mTBI patients have intracranial lesions visible on CT, detecting these lesions is critical for clinical management decisions such as hospitalization and neurosurgical intervention. Existing guidelines focus on CT selection and admission criteria, but precise risk stratification for lesion detection and intervention remains limited. The Stockholm Score aims to fill this gap by leveraging a large multicenter cohort to develop predictive models for lesion identification and clinical outcomes.

Data Highlights

Key Findings

• The pilot cohort included 5,000 adult mTBI patients presenting within 24 hours of injury, all undergoing head CT.
• Intracranial lesions were defined broadly to include hematomas, hemorrhages, diffuse axonal injury, depressed fractures, infarctions, and sinus thrombosis.
• Clinically significant lesions were those leading to neurosurgical intervention, intubation, or death.
• Data collection adhered to TRIPOD guidelines and utilized standardized protocols across three hospitals sharing prehospital and ED management protocols.
• Variables for prediction were selected based on established clinical decision rules and prior literature, including medication use and clinical examination findings.

Clinical Implications

The Stockholm Score has the potential to improve individualized risk assessment for intracranial lesions in mTBI patients, aiding clinicians in decision-making regarding CT imaging and management. Early identification of patients at risk for clinically significant lesions could optimize resource utilization and improve patient outcomes by guiding timely interventions.

Conclusion

This preliminary analysis supports the feasibility of developing robust predictive models for lesion detection in mTBI using the Stockholm Score. Further data collection and analysis will refine these models to enhance clinical utility.

References

1. SELECT-TBI Study Plan and Protocol
2. Scandinavian Neurotrauma Committee Guidelines 2013
3. TRIPOD Statement 2015