RADIOLOGIC-DIAGNOSIS OF TRAUMATIC OCCIPITOVERTEBRAL DISSOCIATION .2. COMPARISON OF 3 METHODS OF DETECTING OCCIPITOVERTEBRAL RELATIONSHIPS ON LATERAL RADIOGRAPHS OF SUPINE SUBJECTS

OBJECTIVE. The purpose of this study was to describe a method for reco gnizing acute traumatic occipitoatlantal dissociation that uses the ba sion-axial interval and basion-dental interval and to compare the accu racy of this method with the accuracies of two other methods: the Powe rs ratio and the x-line method. MATERIALS AND METHODS. Lateral radiogr aphs of the cervical spine of 37 patients in whom the diagnosis of occ ipitoatlantal dissociation had been made on the basis of the relations hip of the basion to the tip of the dens, the Powers ratio, and/or the x-line method were reviewed. Retrospectively, the occipitoatlantal ju nction of each was reassessed by using the Powers ratio, the x-line, a nd the basion-axial interval-basion-dental interval methods. Independe ntly, the neurologic findings at admission and the final neurologic di agnosis at discharge were obtained from the hospital records and were compared with the radiologic findings to determine the degree of radio logic-clinical correlation. RESULTS. Three groups of patients were ide ntified by analyzing the basion-axial and basion-dental intervals of t he occipitovertebral junction and related clinical findings. Twenty-th ree patients (group 1) had frank occipitoatlantal dislocation. Eight p atients (group 2) had incomplete occipitoatlantal dissociation, which was defined as occipitoatlantal subluxation. The remaining six patient s (group 3) had normal radiologic and clinical findings. Four patterns of occipitovertebral dissociation were identified: purely anterior (4 /31, 13%), purely distracted (6/31, 19%), concomitantly anterior and d istracted (20/31, 65%), and purely posterior (1/31, 3%). Regardless of the magnitude or direction of occipitoatlantal dissociation, the basi on-axial interval-basion-dental interval method correctly identified t he abnormality and the type of each. Positive clinical correlation was found in 13 (57%) of the 23 patients in group 1 and in 100% of the ei ght and six patients in groups 2 and 3, respectively. Neither the Powe rs ratio nor the x-line method could be applied in 17 (46%) of 37 case s, either because the opisthion could not be detected on the radiograp hs or because fusion of the posterior arch of C1 had not occurred. In the remaining 20 patients in whom the Powers ratio and the x-line meth od were applicable, the type of occipitoatlantal dissociation was corr ectly identified by the Powers ratio in 12 (60%) and by the x-line met hod in four (20%). Neither the Powers ratio nor the x-line method was applicable in three (50%) of the six patients in whom analysis by the basion-axial interval-basion-dental interval method and clinical findi ngs showed no occipitoatlantal abnormality. Normal occipitovertebral a natomy was correctly identified by the Powers ratio in the remaining t hree patients (50%) and by the x-line method in two (33%). CONCLUSION. Direct measurement of occipitovertebral skeletal relationships altere d by occipitoatlantal dissociation using the basion-axial and basion-d ental intervals provides the most accurate radiologic assessment of th is injury.