Intra- and inter-observer reliability of the novel vertebral bone quality score

To assess the intra- and inter-observer reliability of the novel vertebral bone quality (VBQ) scoring system.

Four orthopedic surgery residents at various levels of training (PGY1-4) evaluated 100 noncontrast, T1-weighted MRIs of the lumbar spine. VBQ was calculated as quotient of the median of L1-L4 average signal intensity (SI) and the L3 cerebral spinal fluid (CSF) SI, as described by Ehresman et al. All measurements were repeated 2 weeks later. We performed a stratified analysis based on patient history of instrumentation, pathology, and MRI manufacturer/magnet strength to determine their effect on VBQ reliability. Spinal pathologies included compression fracture, burst fracture, vertebral osteomyelitis, epidural abscess, or neoplasm. The interclass correlation coefficient (ICC) two-way mixed model on absolute agreement was used to analyze inter-rater and intra-rater reliability. ICC less than 0.40 was considered poor, 0.40-0.59 as fair, 0.60-0.74 as good, and greater than 0.75 as excellent.

Intra-observer reliability was excellent (≥ 0.75) for all four observers. When stratified by history of spinal instrumentation or spinal pathology, all raters showed excellent intra-observer reliability except one (0.71 and 0.69, respectively). When stratified by MRI manufacturer, intra-observer reliability was ≥ 0.75 for all raters. Inter-observer reliability was excellent (0.91) and remained excellent (≥ 0.75) when stratified by history of spinal instrumentation, spinal pathology, or MRI-manufacturer.

VBQ scores from spine lumbar MRIs demonstrate excellent intra-observer and inter-observer reliability. These scores are reliably reproduced in patients regardless of previous instrumentation, spinal pathology, or MRI manufacturer/magnetic field strength.