Is any test 100% specific and 100% sensitive for serious injury?

Abstract

Resnickandcolleagues1haveperformedaprospectiveobservational trial during a 17-month period assessing the accuracy of cervical spine computed tomography (CT) in the evaluation of patients with blunt trauma. The population included patientsata level I traumacenter with persistent cervical tenderness or a focal neurologic deficit. They have also evaluated theusefulnessofperformingmagnetic resonance imaging(MRI) inthesepatients.TheauthorsboldlyconcludethatadvancedgenerationalCTscanning is 100%sensitiveand100%specific indetecting significant cervical spine injuries and thatMRI is of no useexceptmaybe inpatientswith focalmotordeficits.Noneof thepatientsevaluatedinthisstudyhadfocalmotordeficits.This is a very important study because it suggests that MRI is not needed inthispatientpopulation,whichsignificantlysimplifies their care.However, several issuesneedtobeconsideredbefore accepting the authors’ conclusions at face value. Thiswas a very select population ofminimally injured patients with amean Injury Severity Score of 3.3. No patient had anAbbreviated Injury Score greater than 3, andno patient had anInjurySeverityScoregreater than25.Theseresultscannotbe generalized tomoresignificantly injuredpatients.Thevastmajority of patients admitted to level 1 trauma centers would not beincludedinthisstudybecausetheyaremoreseverely injured. The definition of significant injuries was the need for surgical intervention, a halo, or a hard cervical collar for ligamentous injury. Forty-two of 100 MRIs performed showed a ligamentous injury, and the treatment used for these patients is not clear. Five patients with a negative CT who had ligamentous injuries received a cervical collar. However, these patients were not included in the significantly injured group for analysis, and why they were excluded is not clearly presented in the Methods section of the article. Most (61.0%) MRIs were performed as a result of equivocal CT results, and only 26.0%were performed for evaluation of cervical tenderness or focal neurologic findings. This study ismore focusedonequivocalCT findings thanonhowtoevaluate patients with persistent tenderness or a focal neurologic deficit. Finally, theauthorsdonot address theevaluationandcare of patientswithmotor deficits andpotential cervical spine injuries because there were none in the study. This fact remains a significant limitation of the study. Although it would be ideal to dispense with MRI as a diagnostic test in patients with potential cervical spine injuries, this study leavesmanyunansweredquestions. I look forward to more prospective data on this topic from this group of investigators.