Prospective derivation of a clinical decision rule for thoracolumbar spine evaluation after blunt trauma: An American Association for the Surgery of Trauma Multi-Institutional Trials Group Study
Kenji Inaba,Lauren Nosanov,Jay Menaker,Patrick Bosarge,Lashonda Williams,David Turay,Riad Cachecho,Marc de Moya,Marko Bukur,Jordan Carl,Leslie Kobayashi,Stephen Kaminski,Alec Beekley,Mario Gomez,Dimitra Skiada,AAST TL-Spine Multicenter Study Group,Kazuhide Matsushima,Mark Trump,Stephen Varga,Demetrios Demetriades,Thomas Scalea,Diandra Browne,Jeffrey D Kerby,LaDonna Allen,John Berne,Thomas A O'Callaghan,Debra Lillback,Ting Zhao,Douglas Liou,Carlos Brown,Allen Jarzombek,Raul Coimbra,Jennifer Wobig,Muhammad Habib Zubair,Fausto Y Vinces
DOI: https://doi.org/10.1097/TA.0000000000000560
Abstract:Background: Unlike the cervical spine (C-spine), where National Emergency X-Radiography Utilization Study (NEXUS) and the Canadian C-spine Rules can be used, evidence-based thoracolumbar spine (TL-spine) clearance guidelines do not exist. The aim of this study was to develop a clinical decision rule for evaluating the TL-spine after injury. Methods: Adult (≥15 years) blunt trauma patients were prospectively enrolled at 13 US trauma centers (January 2012 to January 2014). Exclusion criteria included the following: C-spine injury with neurologic deficit, preexisting paraplegia/tetraplegia, and unevaluable examination. Remaining evaluable patients underwent TL-spine imaging and were followed up to discharge. The primary end point was a clinically significant TL-spine injury requiring TL-spine orthoses or surgical stabilization. Regression techniques were used to develop a clinical decision rule. Decision rule performance in identifying clinically significant fractures was tested. Results: Of 12,479 patients screened, 3,065 (24.6%) met inclusion criteria (mean [SD] age, 43.5 [19.8] years [range, 15-103 years]; male sex, 66.3%; mean [SD] Injury Severity Score [ISS], 8.8 [7.5]). The majority underwent computed tomography (93.3%), 6.3% only plain films, and 0.2% magnetic resonance imaging exclusively. TL-spine injury was identified in 499 patients (16.3%), of which 264 (8.6%) were clinically significant (29.2% surgery, 70.8% TL-spine orthosis). The majority was AO Type A1 282 (56.5%), followed by 67 (13.4%) A3, 43 (8.6%) B2, and 32 (6.4%) A4 injuries. The predictive ability of clinical examination (pain, midline tenderness, deformity, neurologic deficit), age, and mechanism was examined; positive clinical examination finding resulted in a sensitivity of 78.4% and a specificity of 72.9%. Addition of age of 60 years or older and high-risk mechanism (fall, crush, motor vehicle crash with ejection/rollover, unenclosed vehicle crash, auto vs. pedestrian) increased sensitivity to 98.9% with specificity of 29.0% for clinically significant injuries and 100.0% sensitivity and 27.3% specificity for injuries requiring surgery. Conclusion: Clinical examination alone is insufficient for determining the need for imaging in evaluable patients at risk of TL-spine injury. Addition of age and high-risk mechanism results in a clinical decision-making rule with a sensitivity of 98.9% for clinically significant injuries. Level of evidence: Diagnostic test, level III.