Objective: To develop a simple prognostic score chart for predicting ambulatory status in Thai children with cerebral palsy.

Material and Method: Four hundred seventy one children with cerebral palsy aged 2 to 18 registered and treated at six special schools or hospitals for children with physical disability between 2008 and 2013 were recruited. Baseline characteristics and clinical histories of children with cerebral palsy were collected from medical and physical therapy records. Ambulatory status was classified as three ordinal scales by The Gross Motor Function Classification System - Expanded and Revised version.

Results: Multivariable ordinal continuation ratio logistic regression analysis identified age, type of cerebral palsy, sitting independently at the age of two, and eating independently as significant predictors of ambulation. These items were combined into a clinical prediction score: non-ambulation (scores <7), assisted ambulation (scores 7 to 8), and independent ambulation (scores >8).

Conclusion: The prognostic tool has high discriminative values of ambulatory status among children with cerebral palsy. However, the validation of this tool needs to be tested in other subjects before clinical practice application.

Keywords: Cerebral palsy, Clinical prediction rule, Decision support techniques, Prognosis, Walking