CE: ; /200076; Total nos of Pages: 4; 200076 Repeatability of biomechanical computations based on pelvic radiographic measurements of adult dysplastic hips Blaz ˇ Mavc ˇic ˇ a,b , Marko Kralj a , Vane Antolic ˇ a , Ales ˇ Iglic ˇ c and Veronika Kralj-Iglic ˇ b,c INTRODUCTION A nteroposterior pelvic radiographs are widely used in clinical practice as diagnostic tools for assessment of residual hip dysplasia in adults. 1 Evaluation of the hip joint has been based primarily on morphological measurements of radiographic parameters. With advances in knowledge on hip dysplasia, mathematical models were introduced to complement hip evaluation with estimation of biomechanical parameters. 2,3 Recently, an analytical method HIPSTRESS has been developed that enables computation of the resultant hip force and the contact hip stress distribution from seven measured radiographic parameters on an ante- roposterior pelvic radiograph. 4,5 The method has been used in clinical cross-sectional studies of normal and dysplastic hips. 6,7 Also, by using this method orthopaedic surgical procedures in the hip joint were studied biomechanically. 8,9 If a method is to be used for clinical evaluation of the hip, it should have acceptable repeatability, i.e. pelvic radiographic measurements and biomechanical computations of the same unchanged hip from radiographs taken at different occasions should give the same results. Errors in measurement of pelvic radiographic parameters and computation of biomechanical parameters can arise from unknown magnifications of radio- graphs, 10 different distances of the selected points from the plane of the radiographic film due to variations in the posi- tioning of the pelvis, 11 and the intraobserver/interobserver measurement variability. 12 Studies of other authors on repeat- ability of radiographic measurements in dysplastic hips, which have mainly focused on errors due to different observers 12,13 and errors due to different positioning of the pelvis on differ- ent radiographs, have only been studied on cadavera. 11 There have been no reports on the repeatability of biomechanical computations from such pelvic radiographic measurements. The aim of our paper was to evaluate the repeatability of biomechanical computations based on pelvic radiographic measurements of anteroposterior radiographs of adult dys- plastic hips. In particular, we were interested in determining the repeatability of the resultant hip force computation and the peak contact hip stress computation from repeated radio- graphs of individual hips. MATERIALS AND METHODS Patients The study subjects were selected from among a group of 55 adult patients with hip dysplasia who were treated with a unilateral Bernese periacetabular osteotomy at the ORIGINAL RESEARCH ABSTRACT Background While measurement repeatability of individual pelvic radio- graphic parameters has already been evaluated, there have been no reports on the repeatability of biomechanical computations from such measurements. The aim of our study was to evaluate the repeatability of computations of the resultant hip force and the peak contact hip stress from seven pelvic radiographic parameters in adult dysplastic hips. Methods We performed a retrospective analysis of 18 nonoperated adult dysplastic hips with a series of three anteroposterior pelvic radio- graphs for each hip. The participants were adult patients treated with unilateral Bernese periacetabular osteotomy at our institu- tion in the years 1987–1995, and only their nonoperated hips without osteoarthritis were included. Radiographic measure- ments and biomechanical computations were performed with the previously developed HIPSTRESS method on each of the three radiographs for each individual, and repeatability was assessed with the intraclass correlation coefficients. Results We found high repeatability of the measured Wiberg center-edge angle, pelvic width, the horizontal coordinate of the abductor insertion point on the greater trochanter, intermediate repeat- ability of the interhip distance and low repeatability of the femoral head radius, pelvic height and the vertical coordinate of abductor insertion point on the greater trochanter. The intra- class correlation coefficients of the resultant hip force and the peak contact hip stress were 0.88 and 0.94, respectively. Conclusions The repeatability of biomechanical computations from pelvic radiographic measurements is comparable to the repeatability of the present methods for assessment of hip dysplasia. Keywords hip biomechanics, hip dysplasia, peak contact hip stress, pelvic radiography a Department of Orthopaedic Surgery, University Medical Center Ljubljana, Slovenia b Laboratory of Clinical Biophysics, Faculty of Medicine, University of Ljubljana, Slovenia c Laboratory of Physics, Faculty of Electrical Engineering, University of Ljubljana, Slovenia Correspondence to BlazˇMavcˇicˇ, MD, PhD, Department of Orthopaedic Surgery, University Medical Center Ljubljana, SI-1000 Ljubljana, Slovenia Tel: +386 41 280 385; fax: +386 15 222 474; e-mail: blaz.mavcic@kclj.si 1940-7041 ß 2009 Wolters Kluwer Health | Lippincott Williams & Wilkins Volume 20  Number 3  May/June 2009 Current Orthopaedic Practice 1