Background Lumbosacral transitional vertebra can result in an anomalous number of lumbar vertebrae associated with wrong level treatment. The primary aim of this study was to characterize discrepancies between reported referring levels and levels from MRI reports with treated levels. The secondary aim was to analyze interobserver variability between a pain physician and a radiologist when determining levels and classifying lumbosacral transitional vertebrae.

Methods Between February 2016 and October 2019, a retrospective case series of prospectively collected data of the affected levels mentioned in referrals, MRI reports and treated levels was performed. The counting process, level determination, classification of lumbosacral transitional vertebrae and a secondary control were carried out by independent researchers using a standard methodology.

Results Of the 2443 referrals, 143 patients had an anomalous number of lumbar vertebrae; of these, 114 were included for analysis. The vertebral level noted in the patient’s file, in the referral, and the reported level of treatment differed in 40% of these cases. The vertebral level between the MRI reports and treatment differed in 46% of cases. The interobserver reliability (radiologist vs pain physician) for classifying a transitional vertebra was fair ((κ=0.40) and was substantial (κ=0.70) when counting the vertebrae.

Conclusion In the presence of lumbar spine anomalies, we report a high prevalence of discrepancies between referral levels and MRI pathological findings with treatment levels. Further research is needed to better understand clinical implications.