The utility of magnetic resonance imaging results in physician decision-making before initial lumbar spinal injection

Abstract

BACKGROUND CONTEXTThe need for advanced imaging before spinal intervention is an area of ongoing debate. Many studies have demonstrated the accuracy of magnetic resonance imaging (MRI) results in evaluating structural pathology in the lumbar spine, but few have addressed how frequently MRI findings change clinical management. A randomized controlled trial showed that viewing MRI results did not impact outcomes in patients with radiculopathy undergoing epidural steroid injection (ESI). The results suggested ESIs that correlated with both imaging and clinical findings experienced slightly more benefit than the blinded cohort, although statistically insignificantly. PURPOSEThree related studies were conducted to (1) increase understanding of the opinions of interventional spine physicians regarding the utility of viewing imaging before injection and (2) evaluate the impact of viewing MRI results on injection planning (retrospective and prospective analyses). STUDY DESIGNSurvey, prospective, and retrospective analysis. PATIENT SAMPLEPatients presenting to a university-based spine center for initial evaluation of back or leg pain who were candidates for spinal intervention. OUTCOME MEASURESSelf-reported measures from a clinical practice questionnaire distributed to interventional spine physicians to determine rates and rationale for utilization of MRI before spine injection, physiologic measures including MRI results, functional measures including physician decision-making regarding type and location of injection performed. METHODSThis study was funded by the University of Colorado Health and Welfare Trust. A survey was sent to interventional spine physicians to assess their utilization of MRI results before spine procedures. A retrospective analysis of patients who were candidates for ESI was conducted to evaluate how initial injection plan compared with the postviewing of MRI results on injection performed. In a prospective analysis, injection plans pre- and post-MRI were compared among patients presenting for initial evaluation of low back or leg pain. RESULTSSurvey responses showed that specialists order MRI studies to correlate with physical exam (91%) and to detect the presence of synovial cysts (68%), whereas tumor/infection (93%) was most likely to cause a change in their approach. In the retrospective review, the physician's planned approach before viewing the MRI was concordant with the actual procedure 49% of the time. A different type of procedure was performed in 15% of planned injections. In such cases, the initial treatment plan was altered (ie, same procedure at a different or additional level or side) in 35% of planned injections. In the prospective data collection, 43% of injections were different from the initial physician decision. The most common reasons for altering the injection was different level affected (36%), facet pathology (22%), and different nerve root affected (16%). CONCLUSIONSIn clinical practice, MRI before injection frequently changes management decisions in the planning and delivery of lumbar spine injections.