Classification of spastic hemiplegic cerebral palsy in children

The Winter classification of spastic hemiplegic cerebral palsy (CP) is based on sagittal kinematic data from 3-dimensional gait analysis used in preoperative decision making and postoperative evaluation. Our goal was to investigate how well children with spastic hemiplegic CP can be classified using Winter criteria. Second, we assessed if patients move between groups over time and/or with surgical intervention.

One hundred twelve patients with spastic hemiplegic CP with a mean age of 8.1 years were included. Medical records and the full gait analysis data were reviewed. Patients were classified using Winter criteria, and an independent sample t test was used to compare groups.

We found 26 patients (23%) that could not be classified according to Winter criteria. We defined these patients as group 0. This group showed the least deviation from normal values. Each of the 5 groups in our study showed a higher mean velocity of gait and were younger than any of the groups from the Winter study. In regard to rotational alignment, kinetic variables, and, to a certain extent, muscle tone, group 0 showed the least deviation from normal values; however, most differences were subtle. When reclassifying patients after a mean of 3 years, 8 of 15 had deteriorated in the nonsurgical group, moving to a higher numbered group, whereas 19 of 31 surgically treated patients had improved.

The Winter classification failed to classify 23% (26/112) of our spastic hemiplegic CP children. We suggest that the classification be complemented with the less involved group 0. In this way, all patients can be classified, and thus, treatment plans can be established for all patients. The classification can be divided into ankle, knee, and hip joint involvement. The ankle involvement can be further divided into 3 separate groups. Treating physicians should be aware of the possibility that patients may move into another classification group over time.

Diagnostic level 4. See instructions to authors for a complete description of levels of evidence.