Subclassification of GMFCS level-5 cerebral palsy as a predictor of complications and health-related quality of life after spinal arthrodesis

Background: The Gross Motor Function Classification System (GMFCS) of cerebral palsy categorizes patients by mobility. Patients at GMFCS level 5 are considered the most disabled and at high risk of hip and spine problems, yet they represent a wide spectrum of function. Our aim was to subclassify patients at GMFCS level 5 who underwent spinal arthrodesis on the basis of central neuromotor impairments and to assess whether subclassification predicted postoperative complications and changes in health-related quality of life. Methods: Using a prospective cerebral palsy registry, we identified 199 patients at GMFCS level 5 who underwent spinal arthrodesis from 2008 to 2013. Patients were assigned to subgroups according to preoperative central neuromotor impairments: the presence of a gastrostomy tube, a tracheostomy, history of seizures, and nonverbal status. Nine percent of patients had 0 impairments (GMFCS level 5.0), 14% had 1 impairment (level 5.1), 26% had 2 impairments (level 5.2), and 51% had 3 or 4 impairments (level 5.3). The Caregiver Priorities and Child Health Index of Life with Disabilities (CPCHILD) questionnaire was used for preoperative and postoperative health-related quality-of-life outcome assessments, and major complications were recorded. Results: The rate of major complications increased significantly with higher GMFCS level-5 subtype (p = 0.002), with 12% at level 5.0, 21% at level 5.1, 31% at level 5.2, and 49% at level 5.3. Five of the 7 patients who died within the follow-upperiod were at level 5.3. No significant differences were found among subgroups with respect to the magnitude of correction of the major coronal curve or pelvic obliquity. Preoperative and final follo w-up CPCHILD total scores decreased significantly from GMFCS level 5.0 to level 5.3. However, no significant differences were found by subgroup with respect to the magnitude of improvement in CPCHILD total scores from the preoperative to t he final follow-up evaluation (p = 0.597). Conclusions: Stratification based on central neuromotor impairments can help to identify patients with cerebral palsy at GMFCS level 5 who are at higher risk for developing complications after spinal arthrodesis. Level of Evidence: Prognostic Level III. See Instructions for Authors for a complete description of levels of evidence