Study design: An intra- and interobserver reliability study.
Objective: To assess the intra- and interobserver reliability of the Spinal Deformity Study Group (SDSG) system for classifying adolescent and adult L5-S1 spondylolisthesis.
Summary of background data: Reliability of the SDSG classification has only been previously validated in adolescent patients as performed by the SDSG study group investigators.
Methods: A total of 80 patients with L5-S1 spondylolisthesis were included in this study. Only dysplastic and isthmic spondylolisthesis were included in this study. Long-cassette standing lateral radiographs of the spine and pelvis were obtained. All 80 cases were classified according to the SDSG classification by four observers. After a 2-week interval, the same classification was independently repeated by each observer with the cases in a different randomly assigned order. The Fleiss' κ coefficient was calculated to test the intra- and interobserver reliabilities of the SDSG classification.
Results: The present study included all six types of SDSG classification. Overall intra- and interobserver agreements were 86.6% (κ: 0.830) and 73.3% (κ: 0.648), respectively. The intra- and interobserver agreements and repeatability associated with slip grade were 89.7% (κ: 0.824) and 87.7% (κ: 0.721), respectively. Regarding sacropelvic and spinal balance, intra- and interobserver agreements and repeatability were 83.7% (κ: 0.735) and 77.5% (κ: 0.602) for low-grade slips, and 90.75% (κ: 0.883) and 90.4% (κ: 0.851) for high-grade slips, respectively.
Conclusion: Substantial intra- and interobserver reliability was found for the SDSG classification in L5-S1 lumbar spondylolisthesis. SDSG classification system is a simple and clear classification scheme incorporating spinopelvic parameters, which provides significant clinical utility.
Level of evidence: 3.