Shevell replies

Abstract

SIR-I read with interest and appreciation the thoughtful comments of Rosenbaum et al. The heterogeneity of cerebral palsy (CP) as a symptom complex (in all its ramifications) has been historically recognized and encapsulated in its current consensus definition. 1 It is this heterogeneity that has challenged over time our attempts to capture CP accurately in terms of a single classification scheme. Our paper was an attempt to explore a possible association or relationship between the classical classification scheme based on the distribution and quality of limb impairment and the more recently developed classification scheme based upon gross motor functional capabilities (Gross Motor Function Classification System [GMFCS]). 2 With regard to the specific comments put forward by Rosenbaum et al., I can offer the following responses at the present time.As noted by Rosenbaum et al., we share the view that classification of CP by topographical distribution of impairment is indeed useful for clinical and epidemiological purposes. We also share the opinion that the classification of the severity of gross motor limitations is best captured currently by the GMFCS. In our sample, we found that classification by topographical distribution of motor impairment performed in the child between the age of 2 and 5 years was associated with a dichotomous distinction of GMFCS levels between levels I to III and levels IV to V. This dichotomous distinction was selected because it best approximates the most important functional question posed by parents of young children with a diagnosis of CP: Will my child walk? Detailed examination of the data in our study and that of Gorter et al. reveals remarkable similarities. 2,3 In both, the vast majority of children with hemiplegic or diplegic variants of CP attained a GMFCS level of I to III. Similarly, the majority (roughly 75%) of children with the quadriplegic subtype of CP attained level IV to V. A slight variation was noted in GMFCS levels between the dyskinetic subtypes in the two studies, with roughly 75% attaining level IV to V in our study and 67% in the study of Gorter et al.We would agree with the comments put forward that '… activity limitations are determined only partly by the mere presence of motor impairments'. We never meant to suggest that the distribution of motor impairment is the entire reason for activity limitations. Indeed, other factors must be at play in order to account for the distribution within a subtype amongst GMFCS levels noted in our study and that of Gorter et al. We would also agree that interobserver reliability of neurological subtype assignment has never been objectively evaluated. Indeed, we have reported that neurological subtype may vary over time in a subset (approximately 25%). 4 This variation tends to occur in particular subtypes (i.e. quadriplegic and dyskinetic) and is likely to reflect the influence of central nervous system maturational and contextual factors. This variation in topographical distribution should give clinicians pause in their p...