IMPROVING THE QUALITY OF INTERPRETATION OF PELVIC RADIOGRAPHS FOR HIP SURVEILLANCE IN CEREBRAL PALSY: A QUALITY IMPROVEMENT PROJECT

Hip displacement impacts 35% of children with cerebral palsy (CP) and can result in pain and decreased quality of life. Hip surveillance protocols in CP are designed to facilitate early diagnosis and treatment of hip migration. These include clinical examinations and pelvic radiographs at regular intervals. On pelvic radiographs, hip migration can be measured and risk of progressive hip displacement assessed using Reimer's Migration Percentage (MP). At the outset of this project, issues with hip surveillance at our institution included difficulty in accessing orthopaedic care, inappropriate radiographic technique and inconsistent reporting of MP. The goal of this quality improvement initiative was to improve the technical quality and reporting of hip surveillance radiographs. Change ideas were developed, implemented and evaluated by a multidisciplinary working group, using the Institute for Healthcare Improvement's Model for Improvement framework. Change ideas included development and implementation of a standard radiographic requisition form, as well as an educational campaign focused on reporting of MP. Radiographs were assessed by a paediatric orthopaedic surgeon to determine image quality and MP. Radiographs were analyzed based on the number of views, MP inclusion in radiologist reports, and the concordance of the reported MP/impression between the orthopaedic surgeon and the radiologist. Radiographic quality improved by 33% after the interventions, with 26/51 (51%) deemed good or very good prior to the interventions compared to 54/64 (84%) after the interventions. The number of radiographs judged as poor or very poor decreased by 13%, from 16% (8/51) pre-intervention to 3% (2/64) post-intervention. Concordance between the paediatric orthopaedic surgeon and radiologist for overall hip migration risk increased from 75% (38/51) before the interventions to 92% (59/64) after implementation. The proportion of radiographs that included only two-views increased from 90% (46/51) to 94% (60/64) following the intervention. The number of radiograph reports that included a migration percentage increased from 10% (5/51) to 94% (60/64) following intervention implementation. Implementing an educational campaign and a standardized requisition form for hip surveillance radiographs in children with CP dramatically improved the quality of images and concordance of interpretation between surgeons and radiologists. This will allow ‘decoupling’ of hip surveillance protocols from the orthopaedic surgeon assessment, thereby ensuring that wait times for orthopaedic care do not prevent adherence to hip surveillance protocols. The improved image quality has also facilitated a shift in the institutional protocol from obtaining two views to one view, thereby minimizing radiation exposure for patients.