Purpose of the study: The short- and mid-term symptom-relief of surgical treatment for lumbar stenosis is generally acknowledged, but the probability of a long-term reoperation remains to be studied. The purpose of this work was to determine the long-term risk of reoperation after surgical treatment of degenerative lumbar stenosis and to search for factors influencing this probability.
Material and methods: All patients who underwent from 1989 to 1992 surgical treatment for degenerative lumbar spine stenosis were included in this work. At last follow-up, we noted functional outcome using a specific self-administered questionnaire, patient satisfaction, lumbalgia and radiculalgia using a visual analog scale, SF36 quality-of-life, reoperation or not with time since first operation if performed and the reasons and modalities of the reoperation. The probability of reoperation was determined with the acturarial method. A Cox model was used to search for factors linked with the probability of reoperation; variables studied were: age, comorbid factors, extent of the release, posterolateral arthrodesis or not, extent of the potential fusion, use or not of instrumentation for arthrodesis.
Results and discussion: The study included 262 patients. At last follow-up, 61 patients had died a mean 3.7+/-3 years after the operation; only one of these patients had a second operation 22 months after the first. Forty-four patients were lost to follow-up at mean 6.6+/-3 years. Among these 44 patients, four had a second operation during their initial follow-up at mean 47 months. One hundred fifty-seven patients were retained for this analysis at mean 15+/-1 years follow-up. Among these 157 patients, 29 had a second operation a mean 75 months after the first. There were four reasons for reoperating: insufficient release, destabilization within or above the zone of release, development or renewed zone of stenosis, development or renewed discal herniation. The risk of a second operation was 7.4% [95% CI 4.8-11.6], 15.4% [95% CI 10.7-21.1] and 16.5% [95% CI 11.7-219] at five, 10 and 15 years respectively after the first operation. Among the risk factors studied, only one had a significant impact on reoperation: extent of the zone of release (p=0.003). Compared with a release limited to one level, the risk of reoperation after release of three levels or more was five times greater [95% CI 1.8-12.7].