Predicting functional outcomes following surgical intervention and inpatient rehabilitation stay in children with cerebral palsy

BACKGROUND: The ability to predict functional outcomes in children with cerebral palsy (CP) after surgical procedures and a subsequent inpatient rehabilitation unit (IRU) stay is limited based on current literature. OBJECTIVE: To determine if length of stay (LOS) and functional outcomes could be predicted in children with CP following a surgical procedure and IRU stay based on the surgical procedure performed, pattern of involvement, etiology, and Gross Motor Function Classification System level. DESIGN: Retrospective cohort study. SETTING: Tertiary care pediatrics. PARTICIPANTS: Pediatric patients with CP (N = 139) who underwent one of three surgical procedures followed by an IRU stay. INTERVENTIONS: Selective dorsal rhizotomy, single-event multilevel orthopedic surgery, or intrathecal baclofen pump implantation and subsequent IRU stay. MAIN OUTCOME MEASURES: IRU LOS, Functional Independence Measure for Children (WeeFIM) total score, subscores (mobility, self-care, and cognition), efficiency, and change score. RESULTS: Surgical procedure and WeeFIM mobility subscore at admission were significantly predictive of LOS (p < .001). Type/pattern of limb involvement, age, etiology, Gross Motor Function Classification System, and admission WeeFIM total were significantly predictive of total WeeFIM score at discharge (p < .001) and explained 86% of the variance. All WeeFIM subscores at admission were significant predictors of subscores at discharge (p < .001). Type/pattern of limb involvement, as an individual predictor, was most associated with WeeFIM efficiency (r = .398) and WeeFIM change (r = .445, p < .001). Cognition was found to be a significant mediator of the relationship between both admission mobility and self-care and the gains attained during an IRU stay (p ≤ .004). CONCLUSIONS: LOS and functional outcomes could be predicted with a minimal set of factors accounting for 86% of the variance. Cognition at IRU admission has an important influence on functional outcomes during the postoperative rehabilitation period.