Imaging modalities for low back pain in children: a review of spondyloysis and undiagnosed mechanical back pain.

PURPOSE: Mechanical low back pain is common in the pediatric population; recent studies have shown that undiagnosed mechanical low back pain (UMLBP) is the most common cause of low back pain presenting in adolescents, accounting for up to 78% of cases. Spondylolysis/spondylolisthesis is the most common cause with diagnosed pathology observed in this age group. The goals of this study are to: determine the natural history of low back pain, evaluate the value of radiographic studies in establishing a diagnosis of spondylolysis, and determine the cost and radiation effective doses (EDs) associated with those studies with the associated risks radiation exposure.

METHODS: A retrospective review of patients records aged 10 to 19 years who presented to our institution with mechanical low back of undiagnosed etiology from January 1, 2000 to January 1, 2008 were identified. Patients with previous back surgery, high-energy trauma, congenital syndromes, or medical comorbidities were excluded. UMLBP was defined as back pain with etiology undetected by examination and imaging. We reviewed the following data: age at presentation, sex, the number of follow-up visits, the total length of follow-up, the type of imaging studies performed, and the results from imaging studies. Age-specific radiation EDs were calculated for 10 to 14.9 years, 15 to 18 years, and adults for plain films, fine cut 2-level L-spine computed tomography (CT) scans, and bone scans (BSs).

RESULTS: A total of 2846 patients (63% female) with average age of 14.3 years were identified. A total of 2159 (76%) patients had UMLBP, 61% of that had ≤2 follow-up visits. One hundred and ninety-four patients (7.8%) were diagnosed with spondylolysis; 119 (86%) by plain film, 56 (12.5%) by BSs, and 17 (1.5%) by CTs. Most patients (74%) with spondylolysis had a positive plain film study. There was no significant difference between 2-view (anterior-posterior, lateral) and 4-view (anterior-posterior, lateral, right oblique, left oblique) studies in sensitivity (78% vs. 72%, P=0.39). Advanced imaging was pursued in 90/354 (25%) patients with negative plain film studies. The sensitivity of BS for spondylolysis was 84% (73 of 88 BSs were positive). The sensitivity of CT for spondylolysis was 90% (44 of 49 CTs were positive. BSs exposed patients to much more radiation than CTs and plain film studies.

CONCLUSIONS: Mechanical low back pain is common in adolescents and in most cases is undiagnosed; most require no imaging and ≤2 office visits. For spondylolysis, 2-view plain films are often diagnostic and oblique views did not add significant value. Advanced imaging increases diagnostic accuracy, but adds to the cost and considerable radiation exposure. Because diagnosis of spondylolysis rarely changes clinical management, physicians should use ionizing radiation studies sparingly in children.