Predictors of Ability to Walk after Selective Dorsal Rhizotomy in Children with Cerebral Palsy

SERIAL EVALUATIONS WERE completed after selective dorsal rhizotomy on 90 children with spastic cerebral palsy to analyze whether age, the preoperative gait score, voluntary dorsiflexion at the ankle, the diagnosis (quadriplegia or diplegia), or the length of follow-up correlated with the ability to walk after rhizotomy. The preoperative gait score (P < 0.0001), the diagnosis (diplegia versus quadriplegia, P < 0.0001), unilateral dorsiflexion (P = 0.0029), and bilateral dorsiflexion (P< 0.0001) were significant predictors of the maximal postoperative gait score in the univariate regression analysis, but only the preoperative gait score (P < 0.0001) and the diagnosis (P = 0.0015) retained significant predictive power in the multivariate analysis. These data suggest that the preoperative gait score and the diagnosis are the strongest predictors of ability to walk after selective dorsal rhizotomy. Dorsiflexion demonstrated predictive power only in the univariate model, suggesting that it might have some prognostic value but less than the preoperative gait score or the diagnosis.