Validity and Reliability of Measuring Femoral Anteversion and Neck-Shaft Angle in Patients with Cerebral Palsy

Transcription

1 1195 COPYRIGHT Ó 2010 BY THE JOURNAL OF BONE AND JOINT SURGERY, INCORPORATED Valiity an Reliability of Measuring Femoral Anteversion an Neck-Shaft Angle in Patients with Cerebral Palsy By Chin Youb Chung, MD, Kyoung Min Lee, MD, Moon Seok Park, MD, Sang Hyeong Lee, MD, In Ho Choi, MD, an Tae-Joon Cho, MD Investigation performe at the Department of Orthopaeic Surgery, Seoul National University Bunang Hospital, Kyungki, an the Department of Orthopaeic Surgery, Seoul National University Chilren s Hospital, Seoul, South Korea Backgroun: Increase femoral anteversion an coxa valga are common in patients with cerebral palsy. The purpose of the present stuy was to etermine the valiity an reliability of the methos that are commonly use to measure the proximal femoral geometry in patients with cerebral palsy. Methos: Thirty-six consecutive patients (mean age, eleven years; range, five to twenty years) with cerebral palsy were enrolle in the present stuy. The valiity an the interobserver reliability of the physical examinations performe by three examiners were etermine by comparing the results of a trochanteric prominence angle test, hip internal rotation measurements, an hip external rotation measurements (all with the patient in the prone position) with the amount of femoral anteversion on two-imensional compute tomography. Valiity an intraobserver an interobserver reliability were assesse by comparing the neck-shaft angle on the anteroposterior internal rotation raiograph of the hips with that on the multiplanar reformatte compute tomographic image. Results: The trochanteric prominence angle test showe excellent concurrent valiity (R = 0.862, p < 0.001) an reliability (intraclass correlation coefficient, 0.809). Hip internal rotation also showe goo concurrent valiity (R = 0.787, p < 0.001) an excellent reliability (intraclass correlation coefficient, 0.889), whereas hip external rotation appeare to be unsuitable for preicting femoral anteversion. The neck-shaft angle on the anteroposterior internal rotation raiograph of the hips showe excellent concurrent valiity (R = 0.892, p < 0.001) an reliability (intraclass correlation coefficient, 0.912). Conclusions: A physical examination for etermining femoral anteversion an the neck-shaft angle as measure on the internal rotation raiograph of the hips appear to be clinically relevant methos for evaluating the proximal femoral geometry an version in patients with cerebral palsy. Compute tomographic examination can probably be replace by physical examination an an anteroposterior internal rotation raiograph of the hips for patients with stable hips who are able to walk. Level of Evience: Diagnostic Level I. See Instructions to Authors for a complete escription of levels of evience. Increase femoral anteversion an coxa valga are common proximal femoral conitions in patients with cerebral palsy; these conitions cause an intoeing gait an hip instability, respectively 1-6. Femoral osteotomy is commonly performe to treat these problems in patients with cerebral palsy 7-12, an an important component of surgical planning involves measuring femoral anteversion an the femoral neck-shaft angle. Therefore, femoral anteversion an the neck-shaft angle nee to be measure accurately an reliably to obtain a goo surgical outcome. Several methos have been evise to measure femoral anteversion, such as two-imensional compute tomography 13-16, three-imensional compute tomography 17,18, ultrasonography 19-22, magnetic resonance imaging 23,24, raiography 25-27, an physical examination 28,29. Of these, two-imensional compute tomography measurements have been reporte to Disclosure: The authors i not receive any outsie funing or grants in support of their research for or preparation of this work. Neither they nor a member of their immeiate families receive payments or other benefits or a commitment or agreement to provie such benefits from a commercial entity. J Bone Joint Surg Am. 2010;92: oi: /jbjs.i.00688

2 1196 be accurate 13,30,31, an physical examination is a frequently performe, cost-effective, an safe metho that oes not involve raiation exposure. The physical examination generally inclues a trochanteric prominence angle test 28 along with internal an external hip rotation measurements with the patient in the prone position to assess the rotational profiles. However, there is a concern that the measurements of femoral anteversion in patients with cerebral palsy might be less accurate an reliable than those in normal subjects because spasticity, the abnormal position of the femur 32, an osseous eformities 3,5 can affect a reproucible assessment. The neck-shaft angle of the femur is usually measure on the anteroposterior raiograph of the hips with both hips internally rotate 5,33. However, an inappropriately rotate hip joint on the anteroposterior raiograph can cause an inaccurate measurement of the neck-shaft angle 33, particularly in patients with cerebral palsy. Three-imensional compute tomography as valuable information about the acetabulum when a pelvic osteotomy is being consiere 34,35. However, while accurate 18, this metho may not be necessary for measuring femoral anteversion an the neck-shaft angle only. In the present stuy, it was hypothesize that physical examinations measuring femoral anteversion an the neckshaft angle as measure on an anteroposterior internal rotation raiograph of the hips woul be clinically relevant an woul provie a practical metho for evaluating the proximal femoral geometry in patients with cerebral palsy. Therefore, the present stuy ha two major aims. The first was to examine the valiity an reliability of the trochanteric prominence angle test as well as hip internal rotation an hip external rotation measurements. The secon was to investigate the valiity an reliability of femoral neck-shaft angle measurements on anteroposterior raiographs of the hips. Materials an Methos The present prospective stuy was approve by the institutional review boar at our institution, an informe consent was obtaine from the patients or their guarians. The present stuy inclue consecutive patients with cerebral palsy who were scheule for single-event multilevel surgery between June 2007 an September We inclue patients who were scheule to unergo preoperative compute tomography scans to etermine femoral anteversion an tibial torsion. Patients with a history of orthopaeic intervention (osseous or soft-tissue proceures) for the treatment of cerebral palsy an those with neuromuscular iseases other than cerebral palsy were exclue. Consensus-Builing A consensus-builing session was hel by all examiners before the physical examination an the raiographic an compute tomography measurements of the proximal femoral eformity. Physical examination involving femoral anteversion measurements inclue internal rotation of the hip, external rotation of the hip, an a trochanteric prominence angle test along with the other rotational profiles. All patients were examine in the prone position with the knee flexe to 90. For etermination of the internal rotation angle of the hip, the examiner graspe the patient s ankles an pushe them apart so the leg maximally rotate outwar (Fig. 1). The angle between the vertical line an the long axis of the leg was recore. For etermination of the external rotation angle of the hip, the leg was rotate inwar maximally an the angle between a vertical line an the long axis of the leg was recore (Fig. 2). For the trochanteric prominence angle test 28, with the fingers presse against the greater trochanter, the leg was move into internal or external rotation until the trochanter was palpate in its Fig. 1 With the patient in the prone position, internal rotation of the hip is efine as the angle between a vertical line an the long axis of the leg when the leg is rotate outwar maximally.

3 1197 Fig. 2 With the patient in the prone position, external rotation of the hip is efine as the angle between a vertical line an the long axis of the leg when the leg is rotate inwar maximally. Note that maximum hip external rotation might be inhibite by the contralateral lower extremity. most lateral position, at which time the angle between a vertical line an the long axis of the leg was recore as a physical measurement of femoral anteversion (Fig. 3). When the leg was rotate outwar, the angle was efine as femoral anteversion. When the leg was rotate inwar, the angle was efine as femoral retroversion. All of these measurements were performe by two iniviuals: the examiner (C.Y.C., M.S.P., or S.H.L.) an a physician assistant. Six iniviuals (C.Y.C., M.S.P., S.H.L., an three physician assistants) assesse each patient. Three pairs of orthopaeic surgeons an physician assistants examine the patients consecutively, with each team of examiners being bline to the assessments performe by the other examiners. The examiner hel the rotational position of the leg for each measurement, an the physician assistant measure the angle with use of a stanar universal goniometer with 1 increments an 18-cm-long arms. Care was taken to fix the pelvis firmly uring the physical examination. An anteroposterior raiograph of the hips was mae with the patient in the supine position an with the hips internally rotate approximately 30. The neck-shaft angle of the femur is an angle create by the intersection of a line passing through the miportion of the femoral shaft an another line passing from the femoral hea center through the mipoint of the femoral neck, as measure on the anteroposterior raiograph of the hips with both hips internally rotate (Fig. 4). The center of the femoral hea was efine as the center of the circle of best fit for the femoral hea 36,37. Measuring Valiity an Reliability The valiity of the physical examination with regar to femoral anteversion was assesse by means of correlation with the femoral anteversion as measure on the two-imensional compute tomography scan images (Mx8000-IDT; Philips Fig. 3 The trochanteric prominence angle test measures femoral anteversion, which is efine as the angle between a vertical line an the long axis of the leg when the greatest prominence of the greater trochanter can be palpate laterally.

4 1198 Fig. 4 The neck-shaft angle is measure on the anteroposterior raiograph mae with both hips internally rotate an is efine as the angle create by the intersection of a line passing through the miportion of the femoral shaft an a line passing from the center of the femoral hea through the mipoint of the femoral neck. Healthcare Korea, Seoul, South Korea). Our stanar protocol for a torsional compute tomography scan inclues imaging of the proximal part of the femur, the istal part of the femur, the proximal part of the tibia, an the istal part of the tibia. Femoral anteversion was measure with use of the picture archiving an communication system software (PACS) (IM- PAX; Agfa HealthCare, Mortsel, Belgium) an Rapiia software (version 2.8; Infinitt, Seoul, South Korea) that ha the ability to overlap multiple images. The angle of femoral anteversion was create by a line connecting the posterior margins of each femoral conyle an another line passing through the center of the femoral hea an the mipoint of the femoral neck 13 (Fig. 5). The correlation between the measurements on physical examination an those on two-imensional compute tomography was analyze for concurrent valiity. Specifically, the physical examination measurements mae by one of the examiners were compare with the reference stanar value (i.e., the average of the anteversion measurements mae by two observers on two-imensional compute tomography). It was consiere that this single measurement by a single examiner woul represent the clinical situation more closely. In aition, regression analysis was performe to ientify a linear relationship between the physical examination an compute tomography measurements an possibly to preict a reference value with the physical examination. In the regression analysis, the mean two-imensional compute tomography measurements were compare with the mean physical examination measurements for each patient. The intraobserver an interobserver reliability of the compute tomography measurements of anteversion were assesse with use of the measurements mae by two examiners (M.S.P. an S.H.L.) with eight an seven years of orthopaeic experience, who performe the assessments uring two sessions that were separate by a three-week interval. The interobserver reliability for the measurements of the trochanteric prominence angle test an hip rotation was assesse for three examiners (C.Y.C., M.S.P., an S.H.L.) with twenty-one, eight, an seven years of orthopaeic experience an three physician assistants with more than two years of experience. Before surgery, each examiner performe a phys- Fig. 5 Fig. 6 Fig. 5 Femoral anteversion is measure on the two-imensional compute tomography image as the angle between one line connecting the posterior margins of each femoral conyle an another line passing through the center of the femoral hea an the mipoint of the femoral neck. Fig. 6 With the multiplanar reformation technique, a sectional compute tomography image can be obtaine. On the left, a line connecting the center of the femoral hea an the mipoints of the femoral neck an femoral shaft etermines the sectional image epicte on the right. On the sectional image, the femoral neck-shaft angle is efine as the intersection of a line passing through the miportion of the femoral shaft an a line passing from the center of the femoral hea through the mipoint of the femoral neck.

5 1199 TABLE I Summary of Measurements Examination Mean (eg) Stanar Deviation (eg) Range (eg) Femoral anteversion (two-imensional compute tomography) to 73 Trochanteric prominence angle test to 70 Hip internal rotation to 85 Hip external rotation to 70 Neck-shaft angle on sectional compute tomography to 162 Neck-shaft angle on raiograph to 169 ical examination on a single ay. The examiners were bline to the results of the other examinations. The valiity of the femoral neck-shaft angle as measure on the anteroposterior raiograph of the hips was evaluate by correlating it with the neck-shaft angle on the sectional compute tomography image prouce with the multiplanar reformation technique 38. The DICOM (Digital Imaging an Communications in Meicine) files of the two-imensional torsional compute tomography ata were retrieve, an the proximal part of the femur was reconstructe three-imensionally with use of Rapiia software, which enables the examiner to obtain sectional images from the three-imensional images. On the sectional plane as etermine by the center of the femoral hea, the mipoint of the femoral neck, an the mipoint of the femoral shaft, the neck-shaft angle was measure as the intersection of a line passing through the miportion of the femoral shaft an a line passing from the center of the femoral hea through the mipoint of the femoral neck (Fig. 6). The intraobserver an interobserver reliability of the measurements on the reformatte compute tomography image was assesse with use of the measurements mae by two orthopaeic surgeons, who performe the assessments in two sessions separate by a three-week interval. The intraobserver an interobserver reliability of the femoral neck-shaft angle as measure on the anteroposterior raiograph of the hips was assesse with use of the measurements mae by three examiners. Each examiner measure the neck-shaft angle on the anteroposterior raiograph of the hips twice, with a three-week interval between assessments. The physical examination, raiographic examination, an compute tomography examination were performe within a single week for each patient. All measurements were mae by the examiners without knowlege of the patient s clinical information or the finings of the other examiners. Hip raiographs an compute tomography images were presente to each examiner in ranom orer. Statistical Methos A sample size analysis was carrie out to etermine the minimum number of patients require. In the present stuy, reliability was calculate with use of intraclass correlation coefficients at a target value of 0.8. The 95% confience interval was set to 0.2, an the minimum sample size was thirtysix hips with a Bonett approximation 39. Data from a single hip in each patient were selecte by means of block ranomization an were inclue for statistical analysis. A Kolmogorov-Smirnov test was use to ientify the normality of the variables. Pearson correlation coefficients were use to etermine the concurrent valiity of the physical examinations with regar to femoral anteversion an the neckshaft angle as measure on the anteroposterior raiographs of the hips. In the present stuy, the Pearson correlation coefficient was characterize as poor (0.00 to 0.20), fair (0.21 to 0.40), moerate (0.41 to 0.60), goo (0.61 to 0.80), or excellent (0.81 to 1.00). Linear regression was performe to preict the true femoral anteversion (the two-imensional compute tomography values in the present stuy) from the physical examinations. The intraclass correlation coefficients an their 95% confience intervals were use to summarize the interobserver reliability an were calculate with use of a two-way ranomeffect moel assuming a single measurement an absolute TABLE II Concurrent Valiity of Physical an Raiographic Measurements Examinations R Value P Value Trochanteric prominence angle test an compute tomography measurement <0.001 Hip internal rotation an compute tomography measurement <0.001 Hip external rotation an compute tomography measurement <0.001 Neck-shaft angle on hip raiograph an sectional compute tomography measurement <0.001

6 1200 TABLE III Interobserver Reliability of Physical Examinations Examination Intraclass Correlation Coefficient 95% Confience Interval Trochanteric prominence to angle test Hip internal rotation to Hip external rotation to TABLE IV Intraobserver an Interobserver Reliability of Neck-Shaft Angle as Measure on Hip Raiograph Reliability Intraclass Correlation Coefficient 95% Confience Interval Intraobserver First examiner to Secon examiner to Thir examiner to Interobserver First session to Secon session to Overall to The correlation coefficients between the trochanteric prominence angle test an femoral anteversion as measure on two-imensional compute tomography an between the neck-shaft angle as measure on the anteroposterior raiograph of the hips an the neck-shaft angle as measure on sectional compute tomography were (p < 0.001) an (p < 0.001), respectively (Table II). The interobserver reliability of the physical examination measurements of femoral anteversion was (95% confience interval, to 0.890), (95% confience interval, to 0.937), an (95% confience interval, to 0.702) for the trochanteric prominence angle test, hip internal rotation, an hip external rotation, respectively (Table III). The overall reliability of the neck-shaft angle as measure on the hip raiograph was (95% confience interval, to 0.958) (Table IV). The overall reliability of femoral anteversion as measure on two-imensional compute tomography was (95% confience interval, to 0.990), an that of the femoral neck-shaft angle as measure on the multiplanar reformatte compute tomography image was (95% confience interval, to 0.982) (Table V). With use of regression analysis, the average twoimensional compute tomography measurements an the average physical examination ata for each patient were compare, an the following equations were erive to preict the agreement. An intraclass correlation coefficient of 1 inicates perfect reliability, an an intraclass correlation coefficient >0.8 inicates excellent reliability 40. The level of significance was set at p < Source of Funing There was no external funing source for this investigation. Results Thirty-six consecutive patients (thirty six hips, incluing eighteen right hips an eighteen left hips from a block ranomize selection) with cerebral palsy were enrolle in this stuy. Six patients ha hemiplegia, twenty-five ha iplegia, an five ha quariplegia. The mean age of the patients was 11.0 ± 1.3 years (range, five to twenty years). Twenty-six patients were male, an ten were female. Five, eleven, eleven, seven, an two patients were assigne, respectively, to levels I, II, III, IV, an V accoring to the gross motor functional classification system 41. There were thirty locate hips an six subluxate hips (migration percentage, 33% an <66%) 42. The mean meial proximal tibial angle (an stanar eviation) was 87.9 ± 3.0 (range, 81.8 to 93.9 ). Femoral anteversion as measure on two-imensional compute tomography range from 26 (retroverte) to 73, an the neck-shaft angle as measure on multiplanar reformattecomputetomographyimagesrangefrom118 to 162 (Table I). TABLE V Intraobserver an Interobserver Reliability of Compute Tomography Measurements Reliability Intraclass Correlation Coefficient 95% Confience Interval Femoral anteversion on two-imensional compute tomography Intraobserver First examiner to Secon examiner to Interobserver First session to Secon session to Overall to Neck-shaft angle on multiplanar reformatte sectional compute tomography Intraobserver First examiner to Secon examiner to Interobserver First session to Secon session to Overall to 0.982

7 1201 Fig. 7-A Scatter plot illustrating the relationship between femoral anteversion as measure on two-imensional compute tomography (2DCT) an the trochanteric prominence angle test (TPAT). Note that the regression moel explains the istribution of the ata well, which suggests the clinical relevance of the trochanteric prominence angle test. femoral anteversion as measure on two-imensional compute tomography from the physical examination measurements. Femoral anteversion (on two-imensional compute tomography) = trochanteric prominence angle test (R 2 = 0.829) (Fig. 7-A). Femoral anteversion (on two-imensional compute tomography) = hip internal rotation (R 2 = 0.694) (Fig. 7-B). Femoral anteversion (on two-imensional compute tomography) = hip external rotation (R 2 = 0.472) (Fig. 7-C). Discussion In terms of the physical examination measurements of femoral anteversion, the trochanteric prominence angle test showe excellent valiity an reliability. Hip internal rotation also showe clinically relevant valiity an reliability, whereas hip external rotation i not appear to be a useful inicator of femoral anteversion. The femoral neck-shaft angle measure on the anteroposterior hip internal rotation raiograph also showe goo valiity an excellent reliability. Before iscussing the clinical implications of the present stuy, it is important to aress the limitations of the stuy. First, there might be a ebate regaring the use of the femoral neck-shaft angle on multiplanar reformatte compute tomography images as a reference stanar for the concurrent valiity as previous stuies 13,14,30 have evaluate the accuracy an reliability of femoral anteversion measurements with use of two-imensional compute tomography images. The multiplanar reformatte compute tomography measurements of the neckshaft angle coul reuce the errors resulting from an abnormally positione femur, an it has previously been reporte that the three-imensional compute tomography measurements of the neck-shaft angle were accurate by minimizing the error cause by ifferent positions 18. The accuracy of the measurement of the neck-shaft angle on the multiplanar reformatte compute tomography image has not been confirme, an aitional stuies will be neee even though excellent reliability was shown in the present stuy. Secon, the target population of this stuy was patients with cerebral palsy, an the minimum sample size to present the valiity an reliability of the target population was estimate. The majority of our subjects ha stable hips an were able to walk. Therefore, the factors affecting the valiity an reliability of the femoral geometry measurements, such as the functional level, egree of hip instability, an osseous maturity, coul not be analyze, although six hips were subluxate (migration percentage, 33% an <66%). Care shoul be taken when applying these results to patients with unstable

8 1202 Fig. 7-B Scatter plot illustrating the linear relationship between femoral anteversion on two-imensional compute tomography (2DCT) an hip internal rotation (HipIR). hips who are unable to walk. Thir, our metho for measuring the neck-shaft angle oes not consier an eccentrically locate femoral epiphysis, which is not rare in patients with cerebral palsy. Fourth, in patients with retroverte femora, internally rotating the hips might exaggerate errors in the measurement of the neck-shaft angles on anteroposterior raiographs. There was one retroverte femur in our stuy group, an the measurement error was <5. The physical examination measurements in the present stuy were base on two reference lines, i.e., the longituinal axis of the leg an a line vertical to the floor. In patients with bowe legs, it is ifficult to efine the longituinal axis of the leg. Therefore, this physical measurement coul be istorte in patients with tibia vara or tibia valga 43. However, in the present stuy, the meial proximal tibial angle range from 81.8 to 93.9, an we believe that proximal tibial eformity was not prominent in this stuy group. In patients with severe tibia vara or tibia valga, the longituinal axis shoul be use as a measurement reference for the proximal part of the tibia rather than the whole tibia 44. In the present stuy, the trochanteric prominence angle test showe excellent concurrent valiity an interobserver reliability an appeare to be a clinically relevant metho for measuring femoral anteversion in patients with cerebral palsy. The trochanteric prominence angle test is commonly use because it is believe to be practical, accurate, economical, an safe. However, clinicians shoul be aware that the trochanteric prominence angle test overestimate the compute tomography measurement by a few egrees (mean, 4.8 ± 8.4 ) in the present stuy. The present stuy shows that clinical ata obtaine by means of physical examination were comparable with the compute tomography scan ata. A future stuy will be neee to compare the surgical outcomes for a group of patients evaluate with the trochanteric prominence angle test with those for a group of patients evaluate with compute tomography measurements to etermine the construct valiity of the clinical use of the trochanteric prominence angle test. The egree of hip internal rotation also showe clinically relevant concurrent valiity an interobserver reliability, which suggests that this measurement can be use as an alternative or supplementary metho to the trochanteric prominence angle test when the patient is obese an the trochanter cannot be palpate. The intraclass correlation coefficient between the trochanteric prominence angle test an internal rotation was (p < 0.001). However, the aition of hip internal rotation as another inepenent variable in the regression moel (compute tomography measurement an the trochanteric prominence angle test) cause only a minimal increase in R 2 (0.833). We believe that hip external rotation is an inappropriate metho for preicting femoral anteversion in patients with cerebral palsy.

9 1203 Fig. 7-C Scatter plot illustrating the linear relationship between femoral anteversion on two-imensional compute tomography (2DCT) an hip external rotation (HipER). A possible explanation is that the maximum hip external rotation coul be inhibite by the contralateral limb uring rotation of the teste leg an spasticity of the internal rotator muscles of the hip may prohibit the consistent measurement of external rotation. In a previous stuy, the mipoint of the passive range of hip rotation was reporte to be an inicator of hip rotation uring gait in patients with cerebral palsy 45.However,inthepresent stuy, the valiity an reliability of the trochanteric prominence angle test were superior to the mipoint of maximal hip internal rotation an external rotation (R 2 = 0.730; p < 0.001; intraclass correlation coefficient, 0.755). Ieally, the internal rotation anteroposterior raiograph of the hip shoul be mae with the femoral hea an neck locate perpenicularly to the irection of the raiation in orer to accurately evaluate the femoral neck-shaft angle. However, in practice, hip raiographs are generally mae by meical or parameical personnel, who may not have a knowlege of the amount of femoral anteversion that is present. It may not always be possible to make anteroposterior hip raiographs with internal rotation. Therefore, in the present stuy, all anteroposterior raiographs of the hips were mae with both hips internally rotate by 30 an the neck-shaft angle on these raiographs was compare with the gol stanar (the neck-shaft angle as measure on a compute tomography scan with iniviualize anteversion). In the present stuy, the neck-shaft angle on the anteroposterior internal rotation raiograph of the hips showe clinically relevant reliability, an the mean ifference between the raiographic an compute tomography measurements was 4.0 ± 3.4, with 90% of the measurements being within 10. This fining concurs with those of a previous stuy that emonstrate accurate measurements of the neck-shaft angle on anteroposterior hip raiographs with the femur internally rotate 33.Itis believe that an anteroposterior raiograph of the hips, mae with the hips internally rotate by 30, is a clinically useful metho for etermining the neck-shaft angle of the femur. Several stuies have investigate the valiity an reliability of the trochanteric prominence angle test an the neckshaft angle on raiographs 28,29, However, there have been few stuies on patients with cerebral palsy, even though the proximal femoral eformity is a common problem that can cause hip instability an an abnormal gait an often necessitates surgery in these patients. In the present stuy, measurement of femoral anteversion with use of the trochanteric prominence angle test an measurement of the neck-shaft angle on raiographs appeare to be a clinically relevant metho for evaluating the proximal femoral eformity in patients with cerebral palsy in terms of concurrent valiity an reliability. Thus, a compute tomography examination likely can be replace by a physical examination an an internal

10 1204 rotation anteroposterior raiograph of the hips for patients with stable hips who are able to walk. n NOTE: The authors thank Sung Ju Kim for statistical avice an Mi Seon Ryu for assistance in collecting ata. Chin Youb Chung, MD Kyoung Min Lee, MD Moon Seok Park, MD Sang Hyeong Lee, MD Department of Orthopaeic Surgery, Seoul National University Bunang Hospital, 300 Gumi-Dong, Bunang-Gu, Sungnam, Kyungki , South Korea. aress for M.S. Park: pmsme@hanafos.com In Ho Choi, MD Tae-Joon Cho, MD Department of Orthopaeic Surgery, Seoul National University Chilren s Hospital, 28 Yonkun-Dong, Chongro-Gu, Seoul , South Korea References 1. Laplaza FJ, Root L, Tassanawipas A, Glasser DB. Femoral torsion an neck-shaft angles in cerebral palsy. J Peiatr Orthop. 1993;13: Laplaza FJ, Root L. Femoral anteversion an neck-shaft angles in hip instability in cerebral palsy. J Peiatr Orthop. 1994;14: Bobroff ED, Chambers HG, Sartoris DJ, Wyatt MP, Sutherlan DH. Femoral anteversion an neck-shaft angle in chilren with cerebral palsy. Clin Orthop Relat Res. 1999;364: O Sullivan R, Walsh M, Hewart P, Jenkinson A, Ross LA, O Brien T. Factors associate with internal hip rotation gait in patients with cerebral palsy. J Peiatr Orthop. 2006;26: Robin J, Graham HK, Selber P, Dobson F, Smith K, Baker R. Proximal femoral geometry in cerebral palsy: a population-base cross-sectional stuy. J Bone Joint Surg Br. 2008;90: Crane L. Femoral torsion an its relation to toeing-in an toeing-out. J Bone Joint Surg Am. 1959;41: Amichai T, Harries N, Dvir Z, Patish H, Copeliovitch L. The effects of femoral erotation osteotomy in chilren with cerebral palsy: an evaluation using energy cost an functional mobility. J Peiatr Orthop. 2009;29: Dreher T, WolfS, BraatzF, Patikas D,Döerlein L. Internal rotation gait in spastic iplegia critical consierations for the femoral erotation osteotomy. Gait Posture. 2007;26: Spiegel DA, Flynn JM. Evaluation an treatment of hip ysplasia in cerebral palsy. Orthop Clin North Am. 2006;37:185-96, vi. 10. Hogan KA, Blake M, Gross RH. Subtrochanteric valgus osteotomy for chronically islocate, painful spastic hips. J Bone Joint Surg Am. 2006;88: Ounpuu S, DeLuca P, Davis R, Romness M. Long-term effects of femoral erotation osteotomies: an evaluation using three-imensional gait analysis. J Peiatr Orthop. 2002;22: Settecerri JJ, Karol LA. Effectiveness of femoral varus osteotomy in patients with cerebral palsy. J Peiatr Orthop. 2000;20: Murphy SB, Simon SR, Kijewski PK, Wilkinson RH, Griscom NT. Femoral anteversion. J Bone Joint Surg Am. 1987;69: Sugano N, Noble PC, Kamaric E. A comparison of alternative methos of measuring femoral anteversion. J Comput Assist Tomogr. 1998;22: Weiner DS, Cook AJ, Hoyt WA Jr, Oravec CE. Compute tomography in the measurement of femoral anteversion. Orthopeics. 1978;1: Reikerås O, Høiseth A, Reigsta A, Fönstelien E. Femoral neck angles: a specimen stuy with special regar to bilateral ifferences. Acta Orthop Scan. 1982;53: Kim JS, Park TS, Park SB, Kim JS, Kim IY, Kim SI. Measurement of femoral neck anteversion in 3D. Part 1: 3D imaging metho. Me Biol Eng Comput. 2000;38: Abel MF, Sutherlan DH, Wenger DR, Mubarak SJ. Evaluation of CT scans an 3-D reformatte images for quantitative assessment of the hip. J Peiatr Orthop. 1994;14: Berman L, Mitchell R, Katz D. Ultrasoun assessment of femoral anteversion. A comparison with computerise tomography. J Bone Joint Surg Br. 1987;69: Lausten GS, Jørgensen F, Boesen J. Measurement of anteversion of the femoral neck. Ultrasoun an computerise tomography compare. J Bone Joint Surg Br. 1989;71: Moulton A, Upahyay SS. A irect metho of measuring femoral anteversion using ultrasoun. J Bone Joint Surg Br. 1982;64: Terjesen T, Ana S. Femoral anteversion in chilren measure by ultrasoun. Acta Orthop Scan. 1987;58: Tomczak RJ, Guenther KP, Rieber A, Mergo P, Ros PR, Brambs HJ. MR imaging measurement of the femoral antetorsional angle as a new technique: comparison with CT in chilren an aults. AJR Am J Roentgenol. 1997;168: Guenther KP, Tomczak R, Kessler S, Pfeiffer T, Puhl W. Measurement of femoral anteversion by magnetic resonance imaging evaluation of a new technique in chilren an aolescents. Eur J Raiol. 1995;21: Hermann KL, Egun N. Measuring anteversion in the femoral neck from routine raiographs. Acta Raiol. 1998;39: Kuo TY, Skeros JG, Bloebaum RD. Measurement of femoral anteversion by biplane raiography an compute tomography imaging: comparison with an anatomic reference. Invest Raiol. 2003;38: Lee DY, Lee CK, Cho TJ. A new metho for measurement of femoral anteversion. A comparative stuy with other raiographic methos. Int Orthop. 1992;16: Ruwe PA, Gage JR, Ozonoff MB, DeLuca PA. Clinical etermination of femoral anteversion. A comparison with establishe techniques. J Bone Joint Surg Am. 1992;74: Davis JR, Benfanti P, Blackhurst DW, Allen BL. Assessment of femoral anteversion in chilren with cerebral palsy: accuracy of the trochanteric prominence angle test. J Peiatr Orthop. 2002;22: Mesgarzaeh M, Revesz G, Bonakarpour A. Femoral neck torsion angle measurement by compute tomography. J Comput Assist Tomogr. 1987; 11: Hernanez RJ, Tachjian MO, Poznanski AK, Dias LS. CT etermination of femoral torsion. AJR Am J Roentgenol. 1981;137: Hermann KL, Egun N. CT measurement of anteversion in the femoral neck. The influence of femur positioning. Acta Raiol. 1997;38(4 Pt 1): Kay RM, Jaki KA, Skaggs DL. The effect of femoral rotation on the projecte femoral neck-shaft angle. J Peiatr Orthop. 2000;20: Chung CY, Park MS, Choi IH, Cho TJ, Yoo WJ, Lee KM. Morphometric analysis of acetabular ysplasia in cerebral palsy. J Bone Joint Surg Br. 2006;88: Chung CY, Choi IH, Cho TJ, Yoo WJ, Lee SH, Park MS. Morphometric changes in the acetabulum after Dega osteotomy in patients with cerebral palsy. J Bone Joint Surg Br. 2008;90: Mose K. Methos of measuring in Legg-Calvé-Perthes isease with special regar to the prognosis. Clin Orthop Relat Res. 1980;150: Severin E. Contribution to the knowlege of congenital islocation of the hip. Late results of close reuction an arthrographic stuies of recent cases. Acta Chir Scan. 1941;84(Suppl 63): Sartoris DJ, Resnick D, Bielecki D, Gershuni D, Meyers M. Compute tomography with multiplanar reformation an three-imensional image reconstruction in the preoperative evaluation of ault hip isease. Int Orthop. 1988;12:1-8.

11 Bonett DG. Sample size requirements for estimating intraclass correlations with esire precision. Stat Me. 2002;21: Fleiss JL. Statistical methos for rates an proportions. 2n e. New York: John Wiley an Sons; Palisano R, Rosenbaum P, Walter S, Russell D, Woo E, Galuppi B. Development an reliability of a system to classify gross motor function in chilren with cerebral palsy. Dev Me Chil Neurol. 1997;39: Cobeljić G, Vukasinović Z, Djorić I. Surgical prevention of paralytic islocation of the hip in cerebral palsy. Int Orthop. 1994;18: Gage J, eitor. The treatment of gait problems in cerebral palsy. Lonon: Mac Keith Press; p Lee SH, Chung CY, Park MS, Choi IH, Cho TJ. Tibial torsion in cerebral palsy: valiity an reliability of measurement. Clin Orthop Relat Res. 2009;467: Kerr AM, Kirtley SJ, Hillman SJ, van er Linen ML, Hazlewoo ME, Robb JE. The mi-point of passive hip rotation range is an inicator of hip rotation in gait in cerebral palsy. Gait Posture. 2003;17: Piva SR, Fitzgeral K, Irrgang JJ, Jones S, Hano BR, Brower DA, Chils JD. Reliability of measures of impairments associate with patellofemoral pain synrome. BMC Musculoskelet Disor. 2006;7: Jonson SR, Gross MT. Intraexaminer reliability, interexaminer reliability, an mean values for nine lower extremity skeletal measures in healthy naval mishipmen. J Orthop Sports Phys Ther. 1997;25: Nelitz M, Guenther KP, Gunkel S, Puhl W. Reliability of raiological measurements in the assessment of hip ysplasia in aults. Br J Raiol. 1999;72: Broughton NS, Brougham DI, Cole WG, Menelaus MB. Reliability of raiological measurements in the assessment of the chil s hip. J Bone Joint Surg Br. 1989;71:6-8.