ABSTRACT
BACKGROUND: Migration percentage (MP) is an accepted method of assessing lateral displacement of the femoral head in children with cerebral palsy (CP). Difficulty in positioning of patients for pelvic radiography remains a concern for the reliability of the MP. METHODS: This 2-part quantitative study examined 100 anteroposterior pelvic radiographs for children with CP. Fifty were from a region that had a positioning protocol for hip surveillance of children with CP and 50 images were from a region without. Images were assessed for acceptability of position in relation to hip abduction/adduction and/or pelvic rotation.Ten images deemed Acceptable or Borderline from the region with no protocol were then randomly selected. MP was measured on 2 separate occasions by 5 children's orthopaedic surgeons and statistically analyzed for intrarater and interrater reliability. RESULTS: There was no statistically significant difference in the acceptability of images between the 2 regions with 60% to 66% of the images meeting the criteria outright. When allowances were made for slight variation of abduction/adduction within 5 degrees, 74% to 80% of the images were acceptable.Reliability was variable with limits of agreement between 4.96% and 15.15%. Observers more familiar with the software measuring package had higher reliability within and between occasions. Variability within and between observers decreased as MP increased. CONCLUSIONS: Poor positioning did not appear to be the main reason for the variation in reliability of MP. Repeat measurements were reliable although standardized technique, training, and familiarity with software measuring programmes did influence outcomes.
