RÉSUMÉ
BACKGROUND: Selective lumbar nerve root block (SNRB) is generally accepted as an effective treatment method for back pain with sciatica. However, it requires devices producing radioactive materials such as C-arm fluoroscopy. This study evaluated the usefulness of the longitudinal view of transverse process and needles for medial branch block as landmarks under ultrasonography. METHODS: We performed selective nerve root block for 96 nerve roots in 61 patients under the guidance of ultrasound. A curved probe was used to identify the facet joints and transverse processes. Identifying the lumbar nerve roots under the skin surface and ultrasound landmarks, the cephalad and caudal medial branch blocks were undertaken under the transverse view of sonogram first. A needle for nerve root block was inserted between the two transverse processes under longitudinal view, while estimating the depth with the needle for medial branch block. We then injected 1.0 mL of contrast medium and checked the distribution of the nerve root with C-arm fluoroscopy to evaluate the accuracy. The visual analog scale (VAS) was used to access the clinical results. RESULTS: Seven SNRBs were performed for the L2 nerve root, 15 for L3, 49 for L4, and 25 for L5, respectively. Eighty-six SNRBs (89.5%) showed successful positioning of the needles. We failed in the following cases: 1 case for the L2 nerve root; 2 for L3; 3 for L4; and 4 for L5. The failed needles were positioned at wrong leveled segments in 4 cases and inappropriate place in 6 cases. VAS was improved from 7.6 +/- 0.6 to 3.5 +/- 1.3 after the procedure. CONCLUSIONS: For SNRB in lumbar spine, the transverse processes under longitudinal view as the ultrasound landmark and the needles of medial branch block to the facet joint can be a promising guidance.