Hip rotation in extension and flexion was studied in 23 patients with idiopathic intoeing gait. In extension all the hips had markedly increased medial rotation and limited lateral rotation, fulfilling the criteria of excessive femoral anteversion. In flexion, however, rotation varied widely; in one group of patients medial rotation remained greater than lateral, but in the second group lateral rotation was equal to or greater than medial. CT scans showed that the hips in the first group were significantly more anteverted than those in the second. Clearly measurement of hip rotation in extension alone does not provide a dependable indication of femoral anteversion in children with intoeing gait; rotation in flexion also needs to be measured.

We have considered the reasons for securing containment of the femoral head in Perthes' disease and have reviewed briefly the methods used. The present investigation describes the outcome in a controlled series of forty-eight hips treated by containment by femoral varus-rotation osteotomy in selected patients. In assessing the results we have emphasised that controls are essential, and for this purpose we have used two comparable groups, one untreated and the other treated by methods other than containment. The same factors were considered in assessment--namely age, duration, group, and the presence or absence of "at risk" signs. The results were graded similarly as good, fair and poor in all groups. We have concluded that containment by femoral osteotomy is the treatment of choice in patients with "at risk" signs provided that severe deformity has not already occurred. There is no evidence that treatment of any kind favourably influences the course of the disorder in the remainder. Although this is predominantly a radiological study some clinical features are discussed.