A classification system for hip disease in cerebral palsy

Transcription

1 DEVELOPMENTAL MEDICINE & CHILD NEUROLOGY ORIGINAL ARTICLE A classification system for hip disease in cerebral palsy JONATHAN ROBIN MBBS 1 HKERRGRAHAMMD FRCS (ED) FRACS 1 RICHARD BAKER PHD CENG CSCI 2 PAULO SELBER MD FRACS 3 PAM SIMPSON BSC 4 SEAN SYMONS MBBS FRCS FRCS TR ORTH 3 PAM THOMASON MPT 5 1 Murdoch Childrens Research Institute, The University of Melbourne, Department of Orthopaedic Surgery, The Royal Children's Hospital, Parkville, Victoria, Australia. 2 The Hugh Williamson Gait Laboratory, The Murdoch Childrens Research Institute, Parkville, Victoria, Australia. 3 Department of Orthopaedic Surgery, The Royal Children's Hospital, Parkville, Victoria, Australia. 4 Department of Epidemiology and Preventative Medicine, Monash University, Alfred Hospital, Prahran, Victoria Australia. 5 The Hugh Williamson Gait Laboratory, Department of Orthopaedic Surgery, The Royal Children's Hospital, Parkville, Victoria, Australia. Correspondence to Dr H Kerr Graham at Orthopaedic Department, The Royal Children's Hospital, Flemington Road, Parkville, Victoria 352, Australia. kerr.graham@rch.org.au PUBLICATION DATA Accepted for publication 5th July 28. Published online 3rd December 28. ACKNOWLEDGEMENTS We would like to acknowledge Bill Reid, for all of his work on the illustrations for the Cerebral palsy hip classification. In population-based studies, hip displacement affects approximately one-third of children with cerebral palsy (CP). Given the extreme range of clinical phenotypes in the CP spectrum, it is unsurprising that hip development varies from normality, to dislocation and degenerative arthritis. Numerous radiological indices are available to measure hip displacement in children with CP; however, there is no grading system for assessing hip status in broad categorical terms. This makes it difficult to audit the incidence of hip displacement, determine the relationship between hip displacement and CP subtypes, assess the outcome of intervention studies, and to communicate hip status between health care professionals. We developed a categorical, radiographic classification of hip morphology based on qualitative indices and measurement of the key continuous variable, the migration percentage of Reimers. One hundred and thirty-four radiographs were reviewed of 52 female and 82 male adolescents with CP who were at, or close to, skeletal maturity (mean age 16y 1mo [SD 1y 4mo] range 14y to 19y 1mo). Twenty-nine were classified at Gross Motor Function Classification System level I, 25 at level II, 27 at level III, 24 at level IV, and 29 at level V. A classification system was developed to encapsulate the full spectrum of hip morphology in CP, with and without intervention. Hip displacement is the second most common musculoskeletal deformity affecting children with cerebral palsy (CP), second only to equinus. 1,2 In populationbased studies, hip displacement affects about one-third of children and is directly related to gross motor function as determined by the Gross Motor Function Classification System (GMFCS). 3,4 Hip displacement is also related to the topographical distribution but not to movement disorders, with spastic, dyskinetic, mixed, and hypotonic motor types all having a broadly similar incidence of hip displacement. 3,4 It has been recognized that hip displacement is uncommon in children with spastic hemiplegia, very common in children with spastic quadriplegia, and that the incidence in children with spastic diplegia is in between these two extremes. 1 5 The GMFCS allows gross motor function in children with CP to be described in a valid and reliable manner. 6 Interventions to prevent or manage hip displacement in children with CP vary from injections of botulinum neurotoxin, bracing, phenolization of the obturator nerve, adductor releases, femoral osteotomies, pelvic osteotomies, and salvage procedures. It is not surprising that the interplay between an extremely variable clinical phenotype and variable clinical management results in variable outcomes, ranging from normal hip development through to complete dislocation and degenerative arthritis. Population-based studies are increasingly used in CP, employing CP registers to understand the prevalence of specific clinical problems, such as hip ª The Authors. Journal compilation ª Mac Keith Press 28 DOI: /j x 183

2 displacement, and also to try to understand the outcomes of intervention. In children with developmental dysplasia of the hip (DDH), a grading system known as the Severin classification 7 has been utilized for many years in studies which describe the outcome of management for DDH The classification is based on a number of qualitative radiographic features which describe accurately hip morphology, including the shape of the femoral head, the shape of the acetabulum, the relationship of the femoral head to the acetabulum, as well as directly measured variables such as centre edge angle. 7 The reliability of the Severin classification is enhanced when direct measurement of the centre edge angle is performed as opposed to a qualitative assessment of the overall radiographic features. 15,16 There are obvious differences between typically developing children with DDH and children with CP and spastic hip disease. It is widely accepted that adolescents and young adults with DDH, may have few if any symptoms in the second decade of life. During this period, radiographic abnormalities are an excellent predictor of later degenerative change, symptoms, and the need for joint replacement surgery. We decided to develop a radiographic classification for hips in adolescents with CP, based on gross morphological features and migration percentage of Reimers (MP) around the time of skeletal maturity (typically 14 18y). The goals were to develop a practical morphological classification that could be used to communicate the natural history of hip displacement in children with CP and to describe the outcome of intervention programmes in broad categories. METHOD Qualitative methodology A large, population-based cohort of children with CP born between January 199 and December 1992 has been identified from the Victorian Cerebral Palsy Register and previously reported. 17 Hip development in this cohort has been intensively studied with regular hip radiographs performed, using a standardized protocol, 18 through a hip surveillance clinic and a motion analysis laboratory. We identified a sub-group of 1 adolescents from this cohort who had surgery for hip displacement, and an additional 54 adolescents who did not have any surgery for hip displacement but who had a radiograph around the time of skeletal maturity. Closure of the tri-radiate cartilage is one important sign of skeletal maturation and occurs between 13 to 18 years of age. 19 The lateral acetabular epiphysis, an important secondary growth centre for the acetabulum, first appears and then fuses with the main acetabulum, just before skeletal maturity. 19 Children in this study were included if they had closure of the tri-radiate cartilage of the acetabulum and were aged between 14 and 19 years. This would allow for the most accurate measurement of MP. Radiographs of children who had an open tri-radiate cartilage or were younger than 14 years were excluded as they lacked signs of skeletal maturity, which may have made observations on qualitative and quantitative radiographic features less reliable. The radiographs were studied by the first and second authors for qualitative morphological features that were common to skeletally mature adolescents with CP. The following gross morphological indices were described: (1) integrity of Shenton s arch; (2) shape of the femoral head; (3) shape of the acetabulum; and (4) pelvic obliquity. A classification system was developed utilizing six broad grades anchored at the extremes, with Grade I representing normal hip development and normal radiographic indices, and Grade V signifying complete hip dislocation. As this group included children who had hip surgery, Grade VI was described for those who had a dislocated hip which required salvage surgery. The intermediate grades were developed to reflect important radiographic and prognostic outcomes of hip displacement in children with CP including Grade II a near normal hip, Grade III a dysplastic hip, and Grade IV a subluxated hip. In addition to having a hip which meets all normal radiographic criteria, it was important to have a category describing a near normal hip. It was felt that many treated hips would fall into such a category and in the context of more limited physical demands might continue to function well in adult life. 2 In addition, Grade III was chosen to describe hips with an intermediate level of dysplasia and a poorer outcome from reconstructive surgery 2 into which it was felt many hips would fall. Grade IV was chosen to describe hips with residual subluxation, recognizing that such a hip is unstable, liable to progressive displacement, and symptomatic deterioration with time. 2 It was clearly important to have a category Grade V to describe complete hip dislocation as the incidence of this particular outcome is not well known nor are its associated symptoms. 2 It was also noted that the shape of the femoral head, acetabulum, and degree of pelvic obliquity was more variable in grades IV and V, thus a set of appendices were created to incorporate this variability. The reported incidence of pain in hips that are dislocated varies from 2 to 8%. 21,22 A significant number of dislocated hips are eventually managed by some form of salvage surgery, including valgus osteotomy, 23 resection of the femoral head and part of the proximal femur, 24 combined valgus osteotomy with proximal 184 Developmental Medicine & Child Neurology 29, 51:

3 femoral resection, 25 arthrodesis of the hip, and various forms of replacement arthroplasty. 26 The common feature of all of these forms of salvage surgery is loss of the hip joint. This is rarely considered by patients, parents, or surgeons until the hip is significantly painful. It was important, therefore, to construct a grade encapsulating all forms of salvage surgery as a marker for the incidence of symptomatic hip displacement. Quantitative methodology and validity Following the development of these qualitative grades, quantitative boundaries relating to the hip grade were made using migration percentage. It is commonly accepted that at skeletal maturity the migration percentage for a normal hip is less than 1%. 27 It is also well documented that in CP, hip displacement greater than 3% is referred to as being subluxated and may progress to hip dislocation. 3,28,29 It follows that 1% displacement or more denotes a completely dislocated hip. 27 Grade I and Grades III to V were thus set by these upper and lower limits. The upper confines for Grade II were determined by measuring the migration percentage of many subjectively nearly normal hips. The migration percentage of each hip was measured by the primary author using a standard technique and radiographic reference points. Independently, the second author graded every hip based upon a best estimate of migration percentage and the presence of the qualitative features pertaining to the individual hip grades. The goal of this was to determine face validity by measuring the agreement of the best estimate hip grade and true grade measured by migration percentage. This was thought to test face validity for the classification in a clinical setting where formal measurement of migration percentage for every hip may not be possible. To test concurrent validity of the four qualitative descriptors, both authors collaborated to determine if each morphological qualitative feature matched the true hip grade based on measured migration percentage. Each hip was then given a rating out of four, depending on how many of the qualitative features correlated with the true hip grade. Face validity was measured by using the Kappa statistic with bootstrap confidence intervals and bias was measured using an asymptotic symmetry test. All analyses were performed using Stata 1. (Statacorp, College Station, TX, 27). Concurrent agreement was measured by agreement percentages. Following the results of the comparison of qualitative and quantitative measures of hip displacement, boundaries were confirmed between each of the grades based on migration percentage, creating a hierarchical grading. Combining the quantitative and qualitative parameters, a 6-grade classification system was developed Table I: Distribution according to GMFCS level and motor distribution GMFCS level n Motor distribution n I 29 Spastic hemiplegia 43 II 25 Spastic diplegia 39 III 27 Spastic quadriplegia 52 IV 24 V 29 GMFCS, Gross Motor Function Classification System. encapsulating the full spectrum of hip morphology encountered in a large population of adolescents with CP. RESULTS Demographics The initial cohort consisted of 154 adolescents (38 hips). Of these, 2 were excluded as they were younger than 14 years of age and there was no evidence of tri-radiate cartilage closure at their most recent radiograph. This left radiographs of 268 hips (134 adolescents, 82 males, 52 females) available for analysis. Their distribution according to GMFCS and motor distribution is shown in Table I. Mean age for the group was 16 years 4 months (SD 1y 4mo), range of 14 years to 19 years 1 month. The hip classification The following six grades were constructed for a morphological hip classification for adolescents with CP and are illustrated in Figure 1. Grade I: Normal hip (MP<1%) A morphologically normal hip at skeletal maturity will have a MP <1%. Shenton s arch is intact. The femoral head is round (within 2mm using Mose circles 3 ) and is covered by a well-developed acetabulum. This includes an everted lateral acetabular margin, a normal tear drop, and horizontal sourcil. Pelvic obliquity is <1. Grade II: Nearly normal hip (MP 1 15%) A near normal hip has a MP 1% and 15%. Shenton s arch is intact. The femoral head is round or very close to being round. The acetabulum may be normal or only slightly deformed with the lateral acetabular margin slightly blunted, the tear drop slightly widened but the sourcil should be well formed. Pelvic obliquity is <1. Grade III: Dysplastic hip (MP 16 3%) This is a dysplastic appearing hip. MP is >15% and 3%. Shenton sarchisintact,ormaybebrokenbylessthanor equal to 5mm. The femoral head may be round or slightly flattened. The acetabulum is dysplastic with blunt lateral Cerebral Palsy Hip Classification Jonathan Robin et al. 185

4 Cerebral palsy hip classification Grade I: Normal hip Migration percentage <1% 1. Shenton s arch intact 2. Femoral head round (within 2mm using Mose circles) 3. Acetabulum normal acetabular development with a normal horizontal sourcil, an everted lateral margin and normal tear drop development 4. Pelvic obliquity less than 1 Grade II: Near normal hip Migration percentage 1% 15% 1. Shenton s arch intact 2. Femoral head round or almost round 3. Acetabulum normal or near normal development 4. Pelvic obliquity less than 1 Grade III: Dysplastic hip Migration percentage >15% 3% 1. Shenton s arch intact or broken by less than or equal to 5mm 2. Femoral head round or mildly flattened 3. Acetabulum normal or mildly dysplastic including blunting of the acetabular margin and a widened tear drop 4. Pelvic obliquity less than 1 Grade IV: Subluxated hip Migration percentage >3% <1% 1. Shenton s arch broken by more than 5mm 2. Femoral head variable deformity Appendix I 3. Acetabulum variable deformity Appendix II 4. Pelvic obliquity variable Appendix III Grade V: Dislocated hip Migration percentage 1% 1. Shenton s arch completely disrupted 2. Femoral head variable deformity Appendix I 3. Acetabulum variable deformity Appendix II 4. Pelvic obliquity variable Appendix III Grade VI: Salvage surgery 1. Valgus osteotomy 2. Arthrodesis 3. Excision arthroplasty (Castle) +/ valgus osteotomy (McHale) 4. Replacement arthroplasty Figure 1: Morphological hip classification for adolescents with cerebral palsy. 186 Developmental Medicine & Child Neurology 29, 51:

5 acetabular margins, widened tear drops, and poorly developed sourcil. Pelvic obliquity is <1. Grade IV: Subluxated hip (MP 31 99%) This is a subluxated hip. The MP is >3% but <1%. There is always some contact of the femoral head with the true acetabulum. Shenton s arch is broken by more than 5mm. The femoral head has variable deformity from no deformity to severe deformity with more than half of the femoral head being affected. The acetabulum, likewise, has variable deformity from being normally shaped to having a large gothic arch. There is also variable pelvic obliquity from <1 to >45. To fully grade these hips, reference must be made to Appendices I, II, and III. Grade V: Dislocated hip (MP 1%) This is a dislocated hip. By definition, MP is 1%. There is no remaining contact between the femoral head and true acetabulum. Shenton s arch is completely disrupted. As with Grade IV, there is variable deformity of the femoral head and acetabulum (relating to the timing and rate of dislocation) with variable degrees of pelvic obliquity. To fully grade these hips, reference must be made to Appendices I, II, and III. Grade VI: Salvage surgery, loss of hip joint These hips have undergone salvage surgery because of painful dislocation or subluxation. The salvage surgery technique may be by valgus osteotomy alone, excision arthroplasty 24 plus or minus valgus osteotomy, 25 arthrodesis, or replacement arthroplasty. 26 Validity testing Face validity The agreement between estimated hip grade and true hip grade based on migration percentage was almost perfect, 31 with a Kappa coefficient of.96 (95% confidence interval.93).98). The level of statistical significance was set at 95%. The p value for symmetry test indicated that there was no evidence of bias (p=.32). A validity matrix was constructed and is shown in Figure 2. All disagreements occurred between two adjoining hierarchical grades. True grade Total Estimated grade Concurrent validity Agreement between the four qualitative morphological features and true hip grading by migration percentage are shown in Table II. Almost perfect agreement was displayed for all four qualitative characteristics. 31 Acetabular shape had less agreement with the true hip grade than the other three features. Of all of the disagreements on acetabular shape, 76% were due to the acetabulum being dysplastic (A1 see Appendix I) and 24% were due to worse morphology of a small gothic arch (A2 see Appendix I). All hips with A1 acetabular morphology were either Grade I or II hips (37% and 63% respectively). No Grade I hip had A2 morphology, while only one Grade II hip had an A2 acetabulum. Eighty-three per cent of hips with a small gothic arch were Grade III hips. DISCUSSION In medical research there is an ongoing debate as to whether data are best analyzed by category, reflecting how most physicians think, or continuously, which usually more accurately reflects the data. 32 Categorical data are often useful in the clinical situation such as the outpatient clinic. Continuous data may be more useful in a research setting and are clearly more amenable to sophisticated statistical analysis and interpretation. The ideal situation may be when the categorical data and continuous data are closely related. The best example of direct links between continuous and categorical data in CP is in the assessment and measurement of gross motor function. The Gross Motor Function Measure (GMFM) records continuous data and is the criterion standard tool for valid and reliable measurement of gross motor function in a research setting, 33 but realistically cannot be completed during each clinic visit Total 116 Figure 2: A validity matrix of agreement between estimated hip grade and true hip grade. Table II: Concurrent validity of qualitative hip morphology and true hip grade as measured by migration percentage Qualitative indices Number of hips in agreement (n=268) Shenton s arch Femoral head Acetabulum shape Pelvic obliquity Percentage agreement Cerebral Palsy Hip Classification Jonathan Robin et al. 187

6 The GMFCS is a categorical scale derived from the GMFM and is valid, reliable, and stable over time. 6 It provides an excellent tool for communication between health professionals about gross motor function in children with CP and can be easily recorded at each clinic visit. In CP, there is great heterogeneity in clinical phenotypes and gross motor function, which are directly related to the brain lesion in terms of severity and location. 34 Hip development in CP also follows an extremely variable pattern from normal development to complete dislocation. In our approach to classifying hip morphology in adolescents with CP, we have combined a continuous variable (the MP of Reimers) and developed a categorical classification. Surgeons like categories, because they convey a gestalt about hip function, prognosis, and give strong pointers towards management. However it is important that a categorical classification be valid and reliable. We are gratified, therefore, to find the excellent relationship between the categories of our CP hip classification and the MP of Reimers. MP is the most widely used continuous variable to describe hip displacement in CP. 18 MP is a measurement of the percentage of the femoral head lateral to the acetabular margin. Training and standardizing the measurement technique improves reliability. 18 Given the usefulness of the MP reported in many previous studies, it is reasonable to question whether a categorical classification of CP hip morphology is necessary or useful. First, in hips with significant dysplasia or subluxation, the lateral border of the acetabulum is very difficult to identify accurately, precluding accurate measurement of the migration percentage. In such hips, categorical classification can be much more useful. Second, the constellation of morphological features in each category is suggestive of the prognosis for long-term hip function and may give useful pointers towards management. In this regard, it may be useful to compare the patient s hip status and their GMFCS level. The inclusion of Grade V and VI hips in the classification may be useful in population-based studies concerning the outcome of management of hip disease in CP. The incidence of dislocation in most series with long-term follow-up remains distressingly high, but it is not known with certainty the frequency with which the incidence of symptoms occur in such hips nor the need for salvage surgery. A comparison of Grade V and VI hips should allow a rapid assessment of the incidence of dislocation compared to the incidence of hips which have progressed to salvage surgery. As hip displacement and MP increase, there are clearly observable changes in hip morphology. In hips with normal or very low values of measured migration percentage, the femoral head is round, the acetabulum is well developed, and the pelvis is usually level. With progressive displacement of the femoral head from the acetabulum, variable degrees of femoral head and acetabular deformity are likely to occur, and there is an increasing likelihood of a break in Shenton s arch. In addition, as the lateralized femoral head presses on the lateral margin of the acetabulum, secondary acetabular dysplasia may occur with the development of a dislocation channel or gothic arch. As the femoral head escapes from the acetabulum and is subjected to increased pressure from the spastic and possibly contracted gluteus medius, compression of the femoral head occurs with erosion of cartilage and secondary femoral head deformity. Unilateral hip displacement is frequently associated with pelvic obliquity and scoliosis. In some cases, there is a temporal relationship between unilateral hip displacement, pelvic obliquity, and scoliosis suggesting a causative relationship. In other cases, pelvic obliquity is imposed by the development of a scoliosis before any hip displacement. Although it is readily conceded that radiographic measures of hip morphology are not related automatically to either hip pain, increased disability, or increased burden of care, there is broad agreement in the literature that pain is more common in deformed and dislocated hips than in enlocated hips. 2,22,29 We felt it important to establish a valid classification of hip morphology in CP to describe the full range of hip development from normality through dislocation to the outcomes of salvage surgery. We were gratified to find that each morphologic feature at skeletal maturity was closely associated with the measurement of migration percentage, showing excellent concurrent validity. Further, face validity was displayed with almost perfect agreement between estimated and true hip grades measured with migration percentage, making the classification useful in an everyday clinical setting. Study limitations It must be emphasized that this grading system has been developed around a population-based study with radiographs at or close to skeletal maturity. The final one-third of acetabular development comes from the lateral acetabular epiphysis which does not ossify and join the main acetabulum until just before skeletal maturity. 19 In otherwise normally developing hips, the MP is greater before the appearance of the lateral acetabular epiphysis than afterwards. Therefore, a hip in a 1 to 12-year-old with a MP of 15% may be normal and this migration percentage may spontaneously decrease following the appearance of the lateral acetabular epiphysis prior to skeletal maturity. Hence, this grading system cannot be used reliably in younger children. It remains to be seen whether any adjustments will need to be made in adults with CP. In summary, our classification system provides a 6-level ordinal grading system to describe the full spectrum of hip 188 Developmental Medicine & Child Neurology 29, 51:

7 morphology in adolescents with CP. It offers a means to compare hip development in different populations of adolescents with CP and to describe the outcomes of different management programmes designed to prevent or treat hip displacement. Further work will require reliability testing and validity testing in other age bands. REFERENCES 1. Cornell MS. The hip in cerebral palsy. Dev Med Child Neurol 1995; 37: Graham HK, Selber P. Musculoskeletal aspects of cerebral palsy. J Bone Joint Surg Br 23; 85: Soo B, Howard JJ, Boyd RN, et al. Hip displacement in cerebral palsy. J Bone Joint Surg Am 26; 88: Hägglund G, Lauge-Pederson H, Wagner P. Characteristics of children with hip displacement in cerebral palsy. BMC Musculoskelet Disord 27; 8: Howard CB, McKibbin B, Williams LA, Mackie I. Factors affecting the incidence of hip dislocation in cerebral palsy. J Bone Joint Surg Br 1985; 67: Wood E, Rosenbaum P. The gross motor function classification system for cerebral palsy: a study of reliability and stability over time. Dev Med Child Neurol 2; 42: Severin E. Contribution to the knowledge of congenital dislocation of the hip joint. Acta Chir Scandinavica 1941; 84:(Suppl. 63) Salter RB. Innominate osteotomy in the treatment of congenital dislocation and subluxation of the hip. J Bone Joint Surg Br 1961; 43: Salter RB, Dubos JP. The first fifteen years personal experience with innominate osteotomy in the treatment of congenital dislocation and subluxation of the hip. Clin Orthop 1974; 98: Schoenecker PL, Strecker WB. Congenital dislocation of the hip in children. Comparison of the effects of femoral shortening and of skeletal traction in treatment. J Bone Joint Surg Am 1984; 66: Kasser JR, Bowen JR, MacEwen GD. Varus derotation osteotomy in the treatment of persistent dysplasia in congenital dislocation of the hip. J Bone Joint Surg Am 1985; 67: Zionts LE, MacEwen DG. Treatment of congenital dislocation of the hip in children between the ages of one and three years. J Bone Joint Surg Am 1986; 68: Williamson DM, Benson MKD. Late femoral osteotomy in congenital dislocation of the hip. J Bone Joint Surg Br 1988; 7: Faciszewski T, Keifer GN, Coleman SS. Pemberton osteotomy for residual acetabular dysplasia in children who have congenital dislocation of the hip. J Bone Joint Surg Am 1993; 75: Ward WT, Vogt M, Grucziak JS, Tümer Y, Cook C, Fitch RD. Severin classification system for evaluation of the results of operative treatment of congenital dislocation of the hip. A study of intraobserver and interobserver reliability. J Bone Joint Surg Am 1997; 79: Ali AM, Angliss R, Fujii G, Smith DM, Benson MKD. Reliability of the Severin classification in the assessment of developmental dysplasia of the hip. J Pediatr Orthop Part B 21; 1: Howard J, Soo B, Graham HK, et al. Cerebral palsy in Victoria: motor types, topography and gross motor function. J Paediatr Child Health 25; 41: Parrott J, Boyd RN, Dobson F, et al. Hip displacement in spastic cerebral palsy: repeatability of radiologic measurement. J Pediatr Orthop 22; 22: Tönnis D. Development of the hip joint. In: Tönnis D, Legal H, Graf R. Congenital dysplasia and dislocation of the hip in children and adults. Berlin Heidelberg: Springer-Verlag; 1987, Miller F. Cerebral palsy management hip. Cerebral palsy. New York: Springer Science + Business Media Inc., 25: Andersson C, Mattsson E. Adults with cerebral palsy: a survey describing problems, needs, and resources, with special emphasis on locomotion. Dev Med Child Neurol 21; 43: Bagg MR, Farber J, Miller F. Long-term follow-up of hip subluxation in cerebral palsy patients. J Pediatr Orthop 1993; 13: Hogan KA, Blake M, Gross RH. Subtrochanteric valgus osteotomy for chronically dislocated, painful spastic hips. J Bone Joint Surg Am 27; 89: (Suppl. 2, Pt. 2) Castle ME, Schneider C. Proximal femoral resection-interposition arthroplasty. J Bone Joint Surg Am 1978; 6: McHale KA, Bagg M, Nason SS. Treatment of the chronically dislocated hip in adolescents with cerebral palsy with femoral head resection and subtrochanteric valgus osteotomy. J Pediatr Orthop 199; 1: Root L, Goss JR, Mendes J. The treatment of the painful hip in cerebral palsy by total hip replacement or hip arthrodesis. J Bone Joint Surg Am 1986; 68: Tönnis D. General radiography of the hip joint. In: Tönnis D, Legal H, Graf R. Congenital dysplasia and dislocation of the hip in children and adults. Berlin Heidelberg: Springer-Verlag, 1987, Dobson F, Boyd RN, Parrott J, Nattrass GR, Graham HK. Hip surveillance in children with cerebral palsy. Impact on the surgical management of spastic hip disease. J Bone Joint Surg Br 22; 84: Cooperman DR, Bartucci E, Dietrick E, Millar EA. Hip dislocation in spastic cerebral palsy: long-term consequences. J Pediatr Orthop 1987; 7: Mose K. Methods of measuring Legg-Calvé-Perthes disease with special regard to the prognosis. Clin Orthop 198; 15: Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1977; 33: Bauchner H. Continuous versus categorical data: normalcy versus disease. Arch Dis Child 27; 92: Vos-Vromans DCWM, Ketelaar M, Gorter JW. Responsiveness of evaluative measures for children with cerebral palsy: the gross motor function measure and the Pediatric Evaluation of Disability Inventory. Disabil Rehabil 25; 27: Bax M, Tydeman C, Flodmark O. Clinical and MRI correlates of cerebral palsy. The European Cerebral Palsy Study. JAMA 26; 296: Cerebral Palsy Hip Classification Jonathan Robin et al. 189

8 Appendix I: Femoral head deformity FH: No femoral head deformity A round femoral head with no appreciable deformity. FH1: Mild femoral head deformity A deformity is present, usually on the lateral aspect of the femoral head with loss of cartilage and subchondral bone, extending over less than one quarter of the femoral head. FH2: Moderate femoral head deformity A more extensive deformity of the lateral aspect of the femoral head is present with loss of cartilage and subchondral bone from between one quarter to one half of the femoral head. FH2: Severe femoral head deformity A very extensive deformity of the femoral head is present with the deformity extending over more than half of the femoral head, usually from the lateral margin extending medially. 19 Developmental Medicine & Child Neurology 29, 51:

9 Appendix II: Acetabular deformity A: Normal acetabulum A deep acetabulum with an everted lateral margin, as well as a normal sourcil and tear drop. A1: Dysplastic acetabulum A shallow acetabulum with a blunted lateral margin and thick and widened tear drop. A2: Small gothic arch The lateral margin of the acetabulum is deformed by a gothic arch. A triangular defect is present due to inhibition of ossification or erosion of the lateral margin of the acetabulum from femoral head subluxation. A3: Large gothic arch A large triangular defect is present extending from the pelvic wall towards the mid point of the acetabulum indicating a large ossification defect or erosion. Cerebral Palsy Hip Classification Jonathan Robin et al. 191

10 Appendix III: Pelvic obliquity PO: No pelvic obliquity The pelvis is level or exhibits obliquity of less than or equal to 1 degrees. PO1: Mild pelvic obliquity There is pelvic obliquity of greater than 1 degrees but less than or equal to 25 degrees. PO2: Moderate pelvic obliquity There is pelvic obliquity of greater than 25 degrees but less than or equal to 45 degrees. PO3: Severe pelvic obliquity The pelvic obliquity is greater than 45 degrees. 192 Developmental Medicine & Child Neurology 29, 51: