Background/Purpose: According to different guidelines, the presence of neurological symptoms in patients with lower back pain (LBP) is a validated indication for more complex imaging studies, including lumbar spine magnetic resonance imaging (MRI).  This could help in confirming clinical diagnosis and eventually establishing a treatment plan. We aimed to study the findings and clinical impact of lumbar spine MRI indicated for patients with LBP complicated by neurologic symptoms in our center. 

Methods: We retrospectively analyzed 436 consecutive lumbar spine MRI studies between 2010 and 2013.Our variables of interest were: age, source of the MRI request (general practitioners -GPs- or specialists), “positive” MRI findings (MRI confirming any kind of neurologic compression), correlation between described history and MRI findings, aetiology of neurological compression (osteoarthritis or discal degenerative disease) and finally, the clinical impact on treatment (considered positive for invasive treatment such as steroid injections or surgery versus negative for non-invasive treatment or clinical therapy).

Results: We selected 355 MRIs indicated for LBP with lower limb neurologic symptoms. Patients’ mean age was 52,3 ± 14,1 years, 44.9% female. Most MRI requests, 231(70%), came from GPs. The majority of described neurologic symptoms were sensitive 318 (89,3%). Contrary to our expectations, the rate of normal MRIs was significantly higher in MRIs indicated for motor than sensory symptoms: 15.8% vs 5.7%, respectively (p=0.032). We found normal MRI (with no neurological compressions) in 156 (43.9%) of patients. Among the 199 (56.1%) positive MRIs, there was a dissonance between history and MRI findings in 22 (11,1%) patients. Adding normal MRI studies (156) and MRI with neurologic compressions but without clinical concordance (22) : 178/355 (50,1%) of MRI were negative for clinical relevant neurologic compressions. The likelihood of asking for a positive MRI didn’t differ between GPs and specialists (p=0,84). Only 87 patients (24,4%) had any invasive procedure, 46 (12,9%) had steroid injections and 41 (11,5%) had surgery. Patients whom MRI was prescribed by a specialist instead of a GP had a significantly higher rate of having an invasive procedure: 37.3% vs 18.6%, respectively (p<0.001). 

Conclusion: In this retrospective cohort, half of MRI studies indicated for LBP with supposed neurologic symptoms were negative. In patients with positive MRIs, just one fourth had any kind of invasive procedure. Lumbar spine MRI studies are probably being overused in our center. We hypothesize potential misinterpretation of lower back pain with leg irradiation as true neurologic radicular symptoms.  Moreover, even true radicular pain often resolves with non-invasive treatment. MRI studies should be reserved for refractory cases or those presenting major motor symptoms needing a more invasive treatment such as spinal injections or surgery.

« Back to 2015 ACR/ARHP Annual Meeting

ACR Meeting Abstracts - https://acrabstracts.org/abstract/findings-and-therapeutic-impact-of-magnetic-resonance-imaging-mri-studies-for-patients-with-lower-back-pain-with-neurologic-symptoms-are-we-choosing-wisely/