"Correlation Between Clinical Presentation and Brain CT Findings in Acute Dizziness: A Retrospective Cross-Sectional Analysis at a Tertiary Referral Center"

Background

Dizziness is a frequent presenting complaint in the emergency department (ED), prompting extensive diagnostic evaluation. Non-contrast brain computed tomography (CT) is often utilized to rule out serious central pathologies, but its diagnostic yield is debated, leading to concerns about overuse. This study aimed to identify clinical predictors associated with abnormal brain CT findings in patients with acute dizziness to help refine imaging selection criteria.

Methods

We conducted a retrospective analysis of 291 consecutive adult patients who presented with new-onset dizziness and underwent a non-contrast brain CT scan at Namazi Hospital, a tertiary referral center, between January 2019 and 2021. Patient data, including demographics, comorbidities, clinical symptoms, and hospital outcomes, were extracted from medical records. Statistical analyses were performed to determine associations between clinical variables and CT findings, with odds ratios (OR) and 95% confidence intervals (CI) calculated.

Results

The diagnostic yield of brain CT was low, with a significant majority of scans (72.2%, n=210) revealing no acute pathology. Key clinical factors predicting abnormal CT findings included a history of diabetes mellitus, the presence of ataxic gait, and headache. Conversely, nausea and vomiting were significant predictors of normal findings, being associated with lower odds of central pathology.

Conclusion

The diagnostic yield of routine brain CT in patients with acute dizziness is low. However, specific clinical indicators can effectively stratify risk. The presence of focal neurological signs like ataxia, headache, and certain comorbidities such as diabetes should heighten suspicion for central pathology and support the use of neuroimaging. In contrast, isolated vestibular symptoms like nausea and vomiting are associated with a lower probability of abnormal findings. These results could inform the development of clinical decision rules to optimize CT utilization, thereby reducing unnecessary radiation exposure and healthcare costs.

Key Points

• A large proportion of dizzy patients undergoing non-contrast head CT in the emergency department have no acute radiological abnormalities, underscoring the low overall diagnostic yield of routine CT in this setting.

• In our cohort, diabetes mellitus, ataxic gait, and headache were independently associated with abnormal CT findings, suggesting that these clinical features may help identify patients at higher risk for central pathology.

• Nausea and vomiting were more commonly observed in patients with normal CT scans, consistent with their frequent association with peripheral vestibular syndromes rather than central causes.

• Incorporating simple bedside findings, such as gait assessment and the presence or absence of headache into imaging decisions may help reduce low-yield CT utilization while maintaining attention to high-risk presentations.

• Prospective, multicenter validation of clinically anchored imaging decision rules is needed before these findings can be translated into definitive practice recommendations.