Double trouble - identifying rating inconsistencies due to double ratings of the "Show backbone" study.

Abstract

Background: Currently, the most widely used method to determine lumbar intervertebral disc degeneration is MRI. However, the evaluation of imaging signs of disc degeneration involves several subjective assessments. The aim of this study was to investigate differences in radiological assessments between two independent reports of the same MRI, emphasizing that the identical images were assessed twice by different raters.

Materials: MRI of the lumbar and cervical spine of a population-based random sample of women and a sample of female nurses, geriatric nurses and care workers as a subgroup with a relatively high level of work-related stress on the lumbar spine was performed. Each MRI was then assessed by two radiologists from the corresponding clinic that had examined the participant. Ten criteria were assessed: three continuous and seven categorical parameters. Agreement was assessed with bias and dispersion figures or agreement and Cohen's kappa for categorical parameters.

Results: Double diagnosis of 318 participants with available MR images of the cervical and lumbar spine were performed. The results show that there is remarkable consensus on some parameters as well as substantial disagreement on others-the agreement of the two reports for categorical parameters, as measured by Cohen's kappa, ranges from 0.04 to 0.57. For continuous measurements, the percentage difference ranges from 8 to 24%; it depends on the extent of the subjectivity of the parameter to be rated.

Conclusion: The interrater reliability of MRI readings of the lumbar spine is greater when clearly defined parameters and measurement methods are used. Therefore, it should be investigated which easy to use rating scales can be implemented in daily clinical practice to make reports more reliable and useful for clinicians. One way to reduce subjectivity might be the use of reference images.