Published in:
01-12-2019 | Emergency Radiology
Incidental findings in thoracic CTs performed in trauma patients: an underestimated problem
Published in: European Radiology | Issue 12/2019
Login to get accessAbstract
Objectives
Whole-body CT scans are commonly performed to assess trauma patients, and often reveal incidental findings (IFs) the patient may be unaware of. We assessed the prevalence, associations, and adequacy of follow-up of IFs.
Methods
We retrospectively identified 1113 patients who had a chest CT to assess for traumatic injuries (6-year interval). We coded the radiology reports for IFs and queried our EMR regarding clinical history and adherence to follow-up recommendations for IFs mentioned in the reports.
Results
IFs are much more likely (62.2%) to be found in a chest CT scan than acute traumatic injuries (ATI, 32.4%), in patients being evaluated for potential traumatic injuries. A total of 86.4% of patients who had IFs also had another relevant ICD-10 diagnosis (RD). Lung nodules were the most common IF (45.7%). A multivariate logistic regression model (MLR) demonstrated an accuracy of 89% to predict IFs; the 3 statistically significant predictors (p < 0.05) were any RD (FDR logworth 68.6), followed by smoking history (29.8) and age (4.1). Radiologists recommended follow-up for IF 53.5% of the time, but only 13.9% of patients ever received a follow-up imaging exam or invasive procedure.
Conclusions
IFs are much more common than ATI and can be accurately predicted based on MLR utilizing only 3 clinical variables. While radiologists often recommend follow-up for IFs in trauma patients, most are never effectively followed up or addressed, leading to increased risk of poor outcomes. Clinicians should be aware of the high prevalence of IFs and develop systems for appropriate, evidence-based recommendations, and effective management.
Key Points
• Incidental findings (IFs) are much more common (2×) than acute traumatic injuries (ATI) in chest CTs performed in trauma patients.
• IFs can be accurately predicted via logistic regression modeling with only 3 variables (any relevant ICD-10 diagnosis; positive smoking history; age), which may help radiologist to focus their attention on higher risk patients.
• Radiologists recommend follow-up for IFs more than half of the time; however, IFs are seldom followed up appropriately (less than 14%), leading to missed opportunities and potentially poor patient outcomes.