Interobserver error in interpretation of the radiographs for degeneration of the lumbar spine

We examined 114 segments in 23 patients' lumbar spine plain radiographs affected by disc degeneration. Two consultant orthopaedic surgeons, two consultant radiologists, and one spine nurse practitioner made independent observations on the radiographs. MRI scan films of the corresponding 114 segments were used as a gold standard. Kappa coefficients were used to evaluate the interobserver error, and the error between the independent observers and the MRI scanning reports. The systematic differences between the observers for the diagnosis of the disc degeneration at each segment level was recorded. There was significant interobserver error between the independent observers. The pairwise interobserver agreement ranged from fair to substantial on the plain radiograph observations [Weighted kappa coefficient, mean: 0.517 (CI=0.388-0.646)]. The pairwise interobserver agreement between the independent observers and the MRI scan ranged from fair to moderate [Weighted kappa coefficient, mean: 0.388 (CI=0.259-0.518)]. There is significant error in interpretation of the plain radiographs for the diagnosis of lumbar disc degeneration. MRI may be more accurate in the diagnosis of lumbar disc degeneration.