Ronnie M Davies1, Ashley B Scrimshire2, Lorna Sweetman3, Michael J Anderton2, E Martin Holt2. 1. Department of Orthopaedics, University Hospital of South Manchester, Southmoor Road, Wythenshawe M239LT, United Kingdom. Electronic address: ronnie@etailers.co.uk. 2. Department of Orthopaedics, University Hospital of South Manchester, Southmoor Road, Wythenshawe M239LT, United Kingdom. 3. Christie NHS Foundation Trust, 550 Wilmslow Road, Manchester M20 4BX, United Kingdom.
Abstract
BACKGROUND: Whole-body CT (WBCT) has become routine practice in the assessment of major trauma patients. Whilst this may be associated with increased survival, several studies report high rates of negative scans. As no national guideline exists, selection criteria for WBCT vary widely. This study aims to (1) produce a scoring system that improves patient selection for WBCT (2) quantify patient radiation doses and their concomitant risk of malignancy. METHODS: Clinical notes were reviewed for all patients undergoing a WBCT for trauma over a 21-month period at a UK major trauma centre. Clinical and radiological findings were categorised according to body region. Univariate analysis was performed using Chi-squared testing, followed by multivariable logistic regression. Secondary regression analysis of patients with significant injuries that the model did not identify was performed. The model was optimised and used to develop a scoring system. Sensitivity and specificity were calculated using the same dataset as was used to derive the models. Radiation exposure was determined and the excess lifetime risk of malignancy calculated. RESULTS: 255 patients were included, with a mean age of 45 years. 16% of scans were positive for polytrauma, 42% demonstrated some injury and 42% showed no injury. The regression model identified independent predictors of polytrauma to be (1) clinical signs in more than one body region, (2) reduced Glasgow Coma Score, (3) haemodynamic abnormality, (4) respiratory abnormality, (5) mechanism of injury. The final model had a sensitivity of 95% (95% CI 86-99%) and specificity of 59% (95% CI 52-66%) for significant CT findings. Mean radiation exposure was 31.8 mSv, conferring a median excess malignancy risk of 1 in 474. CONCLUSION: After including neurological deficit, our scoring system had a sensitivity of 97% (95% CI 88-99%) and specificity of 56% (95% CI 49-64%) for significant injury. We propose this is used to stratify the use of trauma radiographs, focused CT and WBCT for major trauma patients. Although not intended to replace clinical judgement, our scoring system adds an objective component to decision-making. We believe this will safely reduce the number of unnecessary CT scans performed on a relatively young cohort of patients.
BACKGROUND: Whole-body CT (WBCT) has become routine practice in the assessment of major traumapatients. Whilst this may be associated with increased survival, several studies report high rates of negative scans. As no national guideline exists, selection criteria for WBCT vary widely. This study aims to (1) produce a scoring system that improves patient selection for WBCT (2) quantify patient radiation doses and their concomitant risk of malignancy. METHODS: Clinical notes were reviewed for all patients undergoing a WBCT for trauma over a 21-month period at a UK major trauma centre. Clinical and radiological findings were categorised according to body region. Univariate analysis was performed using Chi-squared testing, followed by multivariable logistic regression. Secondary regression analysis of patients with significant injuries that the model did not identify was performed. The model was optimised and used to develop a scoring system. Sensitivity and specificity were calculated using the same dataset as was used to derive the models. Radiation exposure was determined and the excess lifetime risk of malignancy calculated. RESULTS: 255 patients were included, with a mean age of 45 years. 16% of scans were positive for polytrauma, 42% demonstrated some injury and 42% showed no injury. The regression model identified independent predictors of polytrauma to be (1) clinical signs in more than one body region, (2) reduced Glasgow Coma Score, (3) haemodynamic abnormality, (4) respiratory abnormality, (5) mechanism of injury. The final model had a sensitivity of 95% (95% CI 86-99%) and specificity of 59% (95% CI 52-66%) for significant CT findings. Mean radiation exposure was 31.8 mSv, conferring a median excess malignancy risk of 1 in 474. CONCLUSION: After including neurological deficit, our scoring system had a sensitivity of 97% (95% CI 88-99%) and specificity of 56% (95% CI 49-64%) for significant injury. We propose this is used to stratify the use of trauma radiographs, focused CT and WBCT for major traumapatients. Although not intended to replace clinical judgement, our scoring system adds an objective component to decision-making. We believe this will safely reduce the number of unnecessary CT scans performed on a relatively young cohort of patients.
Authors: N Moussavi; H Ghani; A Davoodabadi; F Atoof; A Moravveji; S Saidfar; H Talari Journal: Eur J Trauma Emerg Surg Date: 2017-09-25 Impact factor: 3.693
Authors: Kss Dayananda; V Y Kong; J L Bruce; G V Oosthuizen; G L Laing; P Brysiewicz; D L Clarke Journal: Ann R Coll Surg Engl Date: 2018-10-05 Impact factor: 1.891