Study design: Retrospective cohort study.
Objective: To examine health care services use after a magnetic resonance imaging (MRI) scan of the lumbosacral or cervical spine ordered by a primary care physician.
Summary of background data: The use of MRI of the spine in the primary care setting is increasing, yet little is known about the relationship between MRI scan findings and subsequent patterns of health care utilization.
Methods: Linkage of records from an audit of outpatient MRI scans of the spine performed in Ontario, Canada, to administrative databases.
Results: Of the 647 patients who had a lumbosacral spine MRI scan ordered by a primary care physician, 288 (44.5%) were seen in consultation by an orthopedic surgeon or neurosurgeon, and 42 (6.5%) received spine surgery during 3 years of follow-up. Of the 373 patients who had a cervical spine MRI scan ordered by a primary care physician, 164 (44.0%) were seen in consultation by an orthopedic surgeon or neurosurgeon, and none had spine surgery during 3 years of follow-up. Patients with severe disc herniation (likelihood ratio, 5.62, 95% confidence interval, 2.64-12.00) or severe spinal stenosis (likelihood ratio, 2.34; 95% confidence interval, 1.13-4.85) on lumbosacral spine MRI were more likely to undergo subsequent surgery. However, many patients with these MRI abnormalities did not receive surgery, and the absence of these MRI findings did not significantly lower the likelihood of subsequent surgery.
Conclusion: Patients receiving MRI scans of the spine in the primary care setting are frequently referred for surgical assessment and most do not receive subsequent surgery. MRI scan results do not discriminate very well between those who will and will not undergo surgery, suggesting that alternative models for the assessment of patients with spinal complaints in primary care should be explored, particularly in jurisdictions with long wait times for elective spinal surgery consultation.