Towards an Evidence-Based Approach for Imaging in Evaluation of Back Pain in Children

Summary: Patients with back pain often arrive with multiple studies, but an extensive evaluation is negative; such children are said to have "mechanical back pain," (MBP). We evaluated 873 consecutive children between 2003-2004 who presented with back pain, focusing on the 100 patients xrays, MRI, and bone scan. We found that bone scan, not an MRI, is the test of choice to clear patients with nonneurologic back pain of duration less than 6 weeks.

Introduction: Back pain, without neurologic or concerning constitutional symptoms, is one of the most common presenting complaints confronting pediatric orthopaedists. Patients often arrive with multiple studies, but an extensive evaluation is negative; such children are said to have "mechanical back pain," (MBP). We analyzed the value of clinical findings and imaging modalities in reaching a diagnosis in this population.

Methods: We evaluated 873 consecutive children between 2003-2004 who presented with back pain, but no neurologic or other concerning symptoms (fever, etc). We focused on the 100 patients (average age:13 yrs, range 3-17) who had all 3 imaging modalities: lumbar plain radiographs, MRI, and bone scan. We defined Group I as those with all normal radiologic studies (MBP) and Group II as those with an identifiable cause for their back pain. Multiple linear regression and chi-square tests analyzed the predictive capabilities of factors in distinguishing between Groups I and II.

Results: Group I (MBP) had 57 patients, while Group II diagnoses included spondylolysis (n=21), spondylolisthesis (n=6), HNP (n=4), and others (n=12). The ability to detect spondylolysis was superior with bone scan (71%) compared with MRI (57%) and Xray (38%). A negative hyperextension test combined with a negative Xray was highly predictive of Group I (p=0.0085, predictive value of 0.81). The addition of a negative bone scan increased the predictive value to 0.94 (p6 weeks, NPV=0.72, p=0.015).

Discussion: These findings suggest that a bone scan, not an MRI, is the test of choice to clear patients with non-neurologic back pain of duration less than 6 weeks. In detecting spondylolysis, bone scan was superior to MRI which was superior to Xray.