Analysis of the vacuum phenomenon in plain hip radiographs in children.

OBJECTIVES

To find the accurate incidence of the vacuum phenomenon in the hip and the best projection position for producing the vacuum phenomenon in plain hip radiographs in children.

METHODS

All pediatric anteroposterior (AP)-view and frog-leg-position plain hip joint radiographs obtained in our hospital between January 2003 and March 2013 were examined. The subjects' ages ranged between 0 and 14 years (mean, 4.2 years). All of the plain radiographs showing crescent-, linear-, and irregular-shaped lucencies between the femoral head and acetabulum were included in the present study.

RESULTS

A total of 16,749 cases, including 12,422 cases (5,912 boys and 6,510 girls) with only AP-view plain radiographs and 4,327 cases (1,537 boys and 2,790 girls) with both AP-view and frog-leg-position plain radiographs that were assessed in our hospital between January 2003 and March 2013, were examined. None of the AP-view plain hip radiographs exhibited the vacuum phenomenon. Vacuum phenomenon of the hips was found in only 258 cases (321 hips) in the frog-leg-position plain radiographs of 4,327 cases, resulting in a constituent ratio of 5.96% (258/4327). A total of 1,738 normal children were assessed in the 4,327 frog-leg-position radiographs, and 150 cases of the vacuum phenomenon were found in the normal children; thus, the incidence of the vacuum phenomenon in normal children was 8.63% (150/1,738). In 2,360 children with developmental dysplasia of the hip (DDH) who were assessed in the 4,327 frog-leg-position radiographs, 98 cases of vacuum phenomenon were found, yielding an incidence of 4.15% in children with DDH (98/2,360). Thus, the 258 cases with vacuum phenomenon included 150 normal hips (58.14%), 98 cases with DDH (37.98%), 5 cases with Legg-Calvé-Perthes disease (1.94%), and 1 case each of solitary eosinophilic granuloma in the left ischium (0.39%), polyostotic fibrous dysplasia of the left and right proximal femurs (0.39%), 1 case of hereditary multiple exostoses (0.39%), 1 case of congenital coxa vara (0.39%), and 1 fracture of the femoral neck after surgery (0.39%). The 321 hips in the 258 cases were classified as complete (121 hips, 37.69%) or partial (200 hips, 62.31%) types according to the proportion of the lucency area in the hip joints and as linear (159 hips, 49.53%), crescent (151 hips, 47.04%), or irregular (11 hips, 3.43%) lucencies according to the shape of the lucency area in the hips.

CONCLUSIONS

The vacuum phenomenon of the hip in children is found in frog-leg-position plain radiographs. It is easier to find the vacuum phenomenon in normal hips compared to cases with DDH. Frog-leg-position plain radiographs provide a better projection position for obtaining the vacuum phenomenon of the hip in children compared to AP-view plain radiographs.