Sociodemographic variation in use of conservative therapy before MRI of the lumbar spine for low back pain the era of public reporting

Purpose: To evaluate the relationship between use of MRI of the lumbar spine for low back pain without prior conservative therapy and sociodemographic factors after the implementation of public reporting for Medicare's Hospital Outpatient Imaging Efficiency Measure for MRI Lumbar Spine for Low Back Pain (OP-8) metric. Materials and Methods: We conducted a secondary data analysis using a nationally representative sample of 2009 to 2014 Medicare claims to evaluate trends in use of conservative therapy before MRI of the lumbar spine. Continuously enrolled fee-for-service Medicare beneficiaries were included. We applied the same criteria used by Medicare to generate a measure consistent with OP-8. Regression was used to evaluate trends in OP-8 by reporting status (outpatient hospital or clinic) and beneficiary characteristics. Age, sex, and race from the Medicare denominator and area-level socioeconomic measures from the Area Health Resource File were used as covariates. Results: Use of conservative therapy before MRI increased regardless of OP-8 reporting status. Several sociodemographic characteristics were associated with the likelihood of receiving conservative therapy before MRI; beneficiaries were less likely to receive conservative therapy before MRI if they were male, older, black, Hispanic or Latino; if they lived in the West or in an area with more college graduates; or if they had low incomes. Beneficiaries were more likely to receive conservative therapy before MRI if they had poorer health or lived in areas with higher home values. Conclusion: Variations in use of conservative therapy according to factors other than clinically relevant factors, such as health status, are worrying. Further strategies are needed to improve appropriateness and equity in the provision of diagnostic imaging.