Risk factors for detecting significant radiological pathology in patients with acute back injury

Abstract

Introduction: Acute back injury is common in patients presenting to Emergency Departments and radiographic assessment is frequently performed despite the low likelihood of finding a significant injury such as fracture or dislocation. The radiation exposure for thoracolumbar series is high and the low yield of positive pathology suggests that some patients may be exposed to unnecessarily high levels of radiation, associated with prolonged waiting and processing times, possibly with unnecessarily increased cost to the Health Care System. The two objectives of this study were firstly to determine the proportion of significant pathology detected by radiography in the management of patients with back injury, and secondly, to identify useful factors which may be used to discriminate between positive and negative radiological yield. Methods: In a prospective study of patients presenting to the Emergency Department of the Prince of Wales Hospital with acute lower back pain between May and June 1999, standard clinical variables with potential value for discriminating between positive and negative radiological findings in the diagnosis of acute back pain were assessed. Radiographs were reviewed 1) by Emergency physicians, and 2) by two experienced radiologists. The main outcome measure was the presence of a significant finding on radiography, which was defined as any findings warranting further investigation, or follow up. χ2 and Mann-Whitney U tests were used to test whether each variable would discriminate between positive and negative finding on standard lumbo-sacral radiographs. Results: Of 99 patients (60 male) recruited to the study, 45 (45.5%) received lumbo-sacral radiographs and 9 (20%) had significant pathology. Only 2 of 99 patients (2%) received a thoracic radiograph and none of them had significant pathology. The key significant variables identified include: previous back (P=0.08) and bony injury (P=0.04), weight bearing at scene (P=0.003), ability to touch knees (P=0.008) and tenderness over lumbar vertebrae on examination (P=0.007). Of the patients not radiographed, none re-attended with missed injuries. Conclusion: Emergency physicians’ decisions to order thoracolumbar and lumbrosacral radiographs were selective and the yield of positive radiological pathology was low. These findings will be of use in the development of prediction guidelines for requesting radiographs ‘rule in’ and ‘rule out’ significant injury.