Statistical significance of MRI in detection of chance type fractures of thoracolumbar vertebrae

Abstract

BACKGROUND CONTEXT: Chance type fractures, as defined by G Q Chance in 1948, is a flexion-distraction type injury of the spine. There is high suspicion of ligamentous injury on initial X-Rays and CT scan if there is anterior wedging exceeds 40 – 50%, inter-spinous distance widening of more than 20% or vertebral body translation. In case of ligamentous Injury, about half of all patients treated non-operatively will have poor outcomes. Intraoperative finding of ligamentous injury is the gold standard. Although MRI is considered a reliable diagnostic tool for detection of injury to posterior ligamentous complex but limited studies are available that evaluate the correlation statistically. PURPOSE: Quantitatively assess the statistical significance of MRI in detection of suspected Chance type fractures of the thoracolumbar spine. STUDY DESIGN/SETTING: All patients with thoracolumbar vertebral fractures that were operated at HGH, Doha, Qatar from January 2011 to January 2014 were retrospectively assessed. Intraoperative results and reported MRI were used for simple statistical calculations. PATIENT SAMPLE: Non purposive sampling (All eligible patients) OUTCOMEMEASURES: Sensitivity, specificity, positive predictive value, negative predictive value. METHODS: Simple statistical calculations for the Outcome Measures mentioned above were done along with 95% confidence interval calculation. RESULTS: Sensitivity 72.73 % Specificity 100 % PPV 100 % NPV 88.89 % CONCLUSIONS: MRI is not a sensitive indicator for detection of ligamentous injury in Chance type fractures of thoracolumbar vertebra and a positive finding of ligamentous injury on preoperative MRI is highly predictive of injury to the posterior ligamentous complex. CONFLICTS OF INTEREST: None. FUNDING SOURCES: None.