Children with diplegic cerebral palsy develop progressive musculoskeletal deformities with deterioration in their gait. Multilevel surgery is a well-established treatment modality involving a combination of soft tissue lengthening and correction of bony deformities.

At Bristol Royal Children's Hospital we have identified a cohort of 45 children with diplegic cerebral palsy who have undergone multilevel surgery. Video gait analysis had been performed pre-operatively and three years post-operatively. We utilised the Edinburgh Visual Gait Score (EVGS)^[1], a validated system that allows direct comparison with gait videos taken during different periods of the patient's treatment. Seventeen measurements are taken per limb at each stage. The patients were also categorised according to the Functional Walking Score (FWS) ^[2] that assesses their level of independence.

Post-operative results demonstrate a significant improvement in gait score on both the EVGS and FWS. Patients whose gait was more severely affected prior to surgery had the greatest improvement in mobility and functional scores. Patients consistently had significant improvements in hip and knee extension in stance phase, with more modest improvement in knee flexion in swing with persistent co-contraction. Both initial contact and heel lift were consistently abnormal pre-operatively, but few patients achieved a heel strike and normal heel lift post-operatively. We are proceeding with a long-term follow-up of this cohort of patients at 15 years following surgery.

The combination of using detailed video gait analysis with functional assessment is a valuable tool in retrospective assessment of patients' outcome following surgery. It gives a quantitative evaluation of progression over time as well as allowing comparison with a cohort of patients to estimate the future level of functional independence.

Statement of clinical significance: Gait Analysis (GA) is a valuable technique for investigating functional limitations in children with gait abnormalities. Because GA generates such a large quantity of data, it could be more useful to have a single parameter derived from kinematic and kinetic GA data. For this reason, Schutte proposed the use of a global index (Normalcy Index – NI) that is derived from 16 selected gait variables and measures the distance between the patient’s gait data and that of a control population with no pathology. The first aim of this study is the classification of children with gait abnormalities such as “clumsy” children, idiopathic Toe-walkers and children affected by Cerebral Palsy using NI and the second aim is to verify the usefulness of the NI in the characterisation of these subjects’ gait.

Material and methods: The GA trials were carried out at the “Gait Analysis Lab”, Children Hospital “V.Buzzi”, Milan, Italy by using an ELITE system (8 TVC working at 100 Hz) and two force platforms (Kistler, CH). 25 subjects with no known gait pathology (mean age 14, range: 7– 28 years) underwent GA and formed the group needed in order to define the parameters of normal gait. The subjects with gait abnormalities were 7 clumsy children (mean age: 7 years, range: 5-10 years), 17 idiopathic Toe-walkers (mean age: 6 years, range: 5-8 years) and 166 subjects affected by Cerebral Palsy (mean age: 10 years, range: 3-24 years) divided in two groups: Independent Walkers (33 hemiplegics, 106 diplegics and 7 quadriplegics) and Dependent Walkers (13 diplegics and 7 quadriplegics). The mean NI over the available trials was calculated for each subject. For all the subjects the left and right side NI values were pooled. Group means and standard errors were then calculated.

Results: For clumsy children and for idiopathic Toe-walkers we obtained mean NI values higher than mean NI value found for healthy subjects, but they are smaller than the mean NI values calculated for subjects affected by Cerebral Palsy. Moreover for children affected by Cerebral palsy, we found that higher degrees of severity of CP induced impairment were associated with higher NI values, in accordance with the findings of Schutte et al. The division of the Cerebral Palsy subjects into Independent and Dependent Walkers shows that the use of aids results in a locomotor pattern that is totally incomparable with that of “normal gait”.

Conclusions: The NI is easy to understand and to apply in order to summarize GA data. It is a useful element in the classification of the locomotor pattern of subjects with motor abnormalities. The NI is able to distinguish normal subjects from clumsy children and idiopathic Toe-walkers, patients with only minor abnormalities and by using NI it’s possible to classify different levels of functional impairments in group of subjects affected by Cerebral Palsy.