This report examined the surgical treatment results and suggests proper treatment guidelines for revision surgery in the lumbar spine.

This study examined 97 patients, who underwent revision surgery from 1996 to 2004 with a follow-up of at least 1 year. The causes of revision surgery, treatment modalities and results of treatment were analyzed. The clinical results were evaluated using the Kirkaldy-Willis criteria.

In the ninety-seven patients, recurred disc herniation (37 cases, 35 cases: same level, 2 cases: different level) was the leading cause of revision surgery. The other causes were spinal stenosis (17 cases), epidural fibrosis (9 cases), pseudoarthrosis (9 cases), junctional stenosis (9 cases), infection (8 cases), segmental instability (4 cases), metal failure (2 cases) and cyst (2 cases). Spinal fusion was performed in 91 patients, and decompression only was performed in the remaining 6 patients. The clinical results according to reasons for failure were good in cases of recurrent disc herniation. The results of spinal fusion were more successful than that of decompression only (p= 0.002). The results of short segment fusion were more satisfactory than those of long segment fusion (p=0.043). In the final follow-up results, excellent and good results according to Kirkaldy-willis criteria were found in 62 cases.

Spinal arthrodesis is an effective treatment modality for revision surgery in the lumbar spine. However, long segment arthrodesis should be considered carefully because of the high level of disappointing results. In addition, selective arthrodesis of the pathologic level is recommended in revision surgery of the lumbar spine.