COMPARISON OF INTRA- AND INTEROBSERVER RELIABILITY WITH COMPUTED AND CONVENTIONAL RADIOGRAPHY FOR SPINAL DEFORMITY.

Abstract

Children suffering from scoliosis are routinely monitored with spinal radiographs every 4 to 6 months. This radiation exposure represents a concern in the treatment and observation of children with spinal deformity as increased radiation exposure in childhood is known to correlate with increased incidence of both childhood and adult malignancy. Although the examination of spinal deformity is still studied largely by means of conventional radiography, technologic advances have made computed radiography (CR) and digital radiography (DR) more accessible. It has recently been shown that a significant reduction in radiation exposure can be achieved with both CR and DR compared with conventional radiography in the evaluation of children with scoliosis. Although this reduction in radiation exposure does degrade image quality, the interobserver variability in Cobb-angle measurements is not statistically significant. Our institution uses CR in the outpatient evaluation of scoliosis patients. Shriners Hospital for Children, located in Sacramento, uses conventional plain-film radiography. Per the current protocols, both studies result in an essentially equivalent radiation dose to the patient with similar image quality. Because the primary indication for serial studies in scoliosis patients is to follow curve progression, it may be possible to accept a compromised image quality with CR or DR while lowering the radiation dose and maintaining accurate radiologic measurements. The purpose of our study is to compare the intra- and interobserver reliability of clinical measurements made using CR and conventional radiography in the evaluation of children with idiopathic scoliosis, as well as comparing the radiation exposure of each study. Our study involves 25 patients with idiopathic scoliosis between the ages of 8 and 17 seen at Shriners Hospital. The patients are having conventional films taken at Shriners Hospital, followed by CR scoliosis films taken at UC Davis 4 to 6 months later. Four sets of films will then be examined 2 weeks apart by two orthopedic surgeons and two pediatric radiologists. The four film sets are CR films on PACS, CR hardcopy films, conventional films loaded on PACS, and hardcopy conventional films. A variety of clinical measurements will be made and analyzed. The study is currently ongoing. The objective of the study is to measure the accuracy of measurements made using CR compared with conventional radiography. With this established, it may then be possible to lower radiation exposure while maintaining accurate clinical measurements. Ultimately, this would result in improved care of children with scoliosis.