Largest Single Series of Atlanto-occipital Dislocation in Children: Comparison of Radiographic Methods Used in Diagnosis

Summary: A large series of atlanto-occipital dislocations in children was retrospectively reviewed and 3 different radiographic methods (Powers ratio, Kauffman method and Harris method ) used for diagnosis were comparatively evaluated. Dislocations were classified based on horizontal or vertical nature of the displacement, or a combination. An analysis of the three radiographic methods including their specificity and sensitivity is presented with our recommendations.

Purpose: Traumatic atlanto-occipital dislocation (AOD) in children is a rare and often fatal. Plain radiographs and advanced imaging are both used in diagnosis. We present a retrospective review and compare three radiographic methods used for diagnosis.

Methods: We retrospectively identified 16 cases of traumatic AOD in children. Radiographs were reviewed and AOD was classified into anterior displacement (I), vertical distraction (II), posterior displacement (III) and mixed displacement (combination). III and I were considered horizontal dislocation and II was considered a vertical dislocation. 3 established radiological indices were used in each case: Powers ratio, Kauffman method and Harris method. Results were analyzed.

Results: Powers ratio detected 100% (7/7) type I dislocation, 0% (0/4) type III dislocation i.e. 63.6% (7/11) of all horizontal dislocations. Harris et al method detected 85.7% (6/7) type I dislocations, 75% (3/4) type III dislocations i.e. 81.8% (9/11) of all horizontal dislocations. Kauffman method detected 85.7% (12/14) of all type II (vertical) dislocations. By using both Harris and Powers, 90.9% of horizontal dislocations were identified.

Conclusion: Powers ratio is the best indicator of anterior dislocation (100%), but did not identify any cases (0%) of posterior displacement. Harris method was better for detecting anterior displacement (85.7%) than posterior (75%). However, of the methods tested here, Harris was the best for detecting posterior displacement. With combined Powers and Harris method, 90.9% of the cases of horizontal displacement were detected. 85.7% of the cases of vertical displacement were detected by Kauffman method. None of the methods tested here were 100% specific and sensitive for all dislocations. One should note that anatomical landmarks on radiographs may be obscure in young children with incompletely ossified vertebrae. We therefore recommend applying more than one radiographic method for diagnosis of AOD in children. Advance imaging is recommended in all cases of survival.