Explaining the variability improvements in gait quality as a result of single event multi-level surgery in cerebral palsy

Purpose: This is a study of all children with spastic diplegic cerebral palsy (Gross Motor Classification System levels II and III) who had single event multi-level surgery (SEMLS) at a single tertiary referral hospital between 1995 and 2008 to identify factors predicting improvement in gait quality as quantified by the gait profile score (GPS). 9 factors (5 dichotomous and 4 continuous, including preoperative GPS) that might be expected to predict outcomes were identified and univariate and multivariable analysis used to explore how these affected outcomes. Scope: Data from 121 children were included. The mean improvement in GPS of 4.3 degrees was 2.7 times the minimal clinically important difference. Univariate analysis suggested that preoperative GPS is a very strong predictor of improvement in GPS (p < 10(-5)) and when this is considered as a covariate only GMFCS level (p = 10(-5)) and having had previous surgery (p = 0.026) were found to be statistically significant predictors of GPS improvement (p < 0.05). Children of GMFCS level II improved on average by 2 more than those of level III once differences in preoperative GPS had been accounted for. Conclusion: Children with the most abnormal gait patterns preoperatively, and hence those with the most potential to improve are those that improve the most and surgery has clearly been beneficial. Over a quarter of children show changes in GPS which were less than the MCID. The majority of these were those with the least abnormal gait patterns preoperatively and further research is required to establish whether and how such children benefit from SEMLS. (C) 2013 Elsevier B.V. All rights reserved.