Accuracy of migration analysis in hip arthroplasty. Digitized and conventional radiography, compared to radiostereometry in 51 patients.

We assessed the accuracy of migration measurements on conventional and digitized radiographs of total hip arthroplasties by comparing the results with radiostereometry (RSA). 4 stem and 3 acetabular designs were studied. 2 of the stem designs and 1 cup design were uncemented. 180 manual and 202 digitized measurements were done on 120 conventional radiographic examinations. The readings on digitized radiographs did not differ from the manual measurements on the same radiographs. A comparison of the measurements from plain radiographs and with RSA of the femoral stems revealed an accuracy (absolute mean + 2 SD) of between 3.9 and 12.3 mm, depending on the choice of landmarks. The greatest accuracy was obtained by using tantalum markers inserted into the greater or the lesser trochanter and the shoulder of the stem. The most medial point of the lesser trochanter proved to be the best bony landmark. Measurements, including both the center of the femoral head and the greater trochanter, were associated with poor accuracy. The accuracy as regards horizontal cup migration varied from 4.4 to 6.5 mm and the accuracy as regards vertical migration varied between 4.4 and 6.3 mm. The intraobserver error (2 SD) ranged from 1.6 mm to 5.6 mm, The corresponding figures for the inter-observer error were 2.6 mm and 6.6 mm, respectively. One of the cemented cup designs was associated with inferior accuracy. Stem design did not affect the accuracy.