A classification for hip disease in cerebral palsy

Abstract

hip disorders in patients with CP was equally poor, with some research stating that up to 50% of patients with ‘dislocating’ hips developed hip pain when the patient aged. Who were these patients and what were their hip disorders? Reimers, in his classic work on the stability of the hip in children, added the migration percentage of hip subluxation to the lexicon of classification systems. 2 While this has become the standard of hip measurement in CP, it is still difficult to determine where in the spectrum of disease a particular patient lies. Reimers’ classification also does not address the degree of acetabular involvement. Why do classification systems matter? CP is a diverse spectrum of disorders ranging from mild motor disorder (GMFCS level I) to severe total body involvement (GMFCS level V). Similarly, mild hip disorders with minimal structural changes need not have any treatment while severe hip dislocation with femoral head destruction and acetabular involvement may adversely affect the quality of life of a patient who has such a hip. Comparing large groups of patients who have similar, measureable conditions (in this case hip dysplasia) enables the researcher and the clinician to understand the natural history of the problem and, perhaps, make conclusions about the best way to treat their patients. Treatments such as physical therapy, chemical denervation, bracing, orthopedic surgery, etc. purport to ‘improve’ the hip, yet it is often difficult to determine the degree of involvement that any particular patient had prior to or after the treatment. A classification system that has discrete parameters enables one to ‘put that disorder into a box’ for communication with other clinicians and other researchers. While I applaud this attempt to categorize hip problems in CP, there are a few limitations of the Robin et al. study. First, although the authors performed validity testing, there are other studies demonstrating the inter- and intrarater error in using the Reimers’ index. 3 Second, the quality of the radiographs, i.e. standardization of positioning of the hip was not discussed. The measurements become more difficult and less reliable with pelvic obliquity and rotation. Third, this is a classification on skeletally mature patients and it may be too late to intervene to improve hip function by the time that the patient is skeletally mature and may lessen the impact of a classification system. Last, hip disorders in CP are a three-dimensional problem. Often the plain radiograph appears fairly normal, yet the hip may be subluxated posteriorly. Clearly three-dimensional imaging (computed tomography or magnetic resonance imaging) would add to the classification of hip disorders in children with CP. 4 However, with these caveats, the authors should be congratulated for taking a complex problem and attempting to break it down to discrete categories. This, I believe, is one of the challenges of science: taking continuous data that is overwhelming and attempting to organize it for our human brains (especially orthopedic surgeon’s brains). Combining this hip classification schema with the GMFCS will enable us to evaluate our patients in the clinic and in our clinical research to determine if any of our therapies change the natural history of the hip in children with CP. Why another classification system? It’s all about communication.