STUDY DESIGN: A register cohort study.
OBJECTIVE: To evaluate outcome of surgery for lumbar spinal stenosis without concomitant degenerative spondylolisthesis according to predominance of pain and to analyze the role of spinal fusion in conjunction with decompression in patients with predominant back pain (BP) or leg pain (LP).
SUMMARY OF BACKGROUND DATA: Predominance of BP is associated with inferior outcome of surgery for central spinal stenosis. It is unknown if using spinal fusion improves outcomes.
METHODS: In a register study of 9051 patients, we studied outcome of surgery in terms of BP and LP visual analogue scale, function (the Oswestry Disability Index and self-estimated walking distance), health-related quality of life (Short-Form 36 and EuroQol), and patient satisfaction. Outcome was analyzed for 4 groups at 1- and 2-year follow-ups; preoperative BP was equal to or worse than LP and decompression, preoperative BP was equal to or worse than LP and decompression and fusion, preoperative BP was less than LP and decompression, preoperative BP was less than decompression and fusion.
RESULTS: Patients with concomitant fusion were younger and had higher BP and Oswestry Disability Index scores and lower preoperative EuroQol. Predominant BP was associated with inferior outcome in terms of pain, health-related quality of life, and function. Patients most often satisfied (69%) were patients with BP less than LP treated with decompression and fusion and the least satisfied group was patients with BP equal to or worse than LP treated with decompression (54%). Fusion was not only associated with higher EuroQol at 1-year follow-up for patients with predominant BP, but also associated with increased LP at 2-year follow-up in patients with predominant LP. Patients with predominant BP experienced small gains in the physical component summary with fusion.
CONCLUSION: Predominance of BP is associated with inferior outcome. Using spinal fusion improves unadjusted outcome but the benefit is small and not clinically significant and generally disappears in the adjusted analysis.
LEVEL OF EVIDENCE: 4.
Lippincott Williams & Wilkins, 2014. Vol. 39, no 3, p. E199-E210