23. Outcomes of direct vs indirect decompression for lumbar spondylolisthesis: A propensity matched cohort analysis

BACKGROUND CONTEXT Debate persists regarding the optimal surgical strategy to treat lumbar spondylolisthesis. The recent development of novel anterior approaches (direct lateral and oblique lumbar interbody fusion) has led to a rise in the utilization of anterior lumbar interbody fusion procedural codes. These new techniques rely on indirect decompression to treat the patient's neural compression, which is criticized by proponents of direct decompressive techniques. PURPOSE To compare 3- and 12-month outcomes between patients with lumbar spondylolisthesis treated with direct decompression (DD) vs indirect decompression (ID) and interbody fusion. STUDY DESIGN/SETTING Lumbar module of the Quality Outcomes Database (QOD), a national, multicenter prospective spine registry. PATIENT SAMPLE A total of 4,163 patients underwent DD (posterior lumbar laminectomy with interbody fusion) and 86 patients underwent ID (anterior lumbar interbody fusion and posterior instrumentation/fusion without laminectomy). OUTCOME MEASURES Oswestry Disability Index (ODI), NRS-back pain, NRS-leg pain, EQ5D, satisfaction, major complication, minor complication, readmission, and return to work at 3- and 12-months postsurgery, return to OR within 30 days. METHODS Propensity scores were estimated using logistic regression with surgical procedure as the dependent variable. Independent variables included the baseline covariates that are potentially associated with outcomes. Based on the logit of estimated propensity score, DD patients were one-to-one matched to ID patients using a greedy matching strategy without replacement. Binary outcomes between ID and DD propensity score matched patients were compared using McNemar's test. The estimates of risk difference and relative risk were determined as well as 95% confidence intervals using asymptotic score methods. Continuous outcomes were compared using two sample paired t-test. RESULTS In the propensity-matched analysis, the DD and ID cohorts did not differ with respect to their 12 month postoperative improvement in ODI (p=0.262), back pain (p=0.715), leg pain (p=0.144), EQ5D (0.384), satisfaction (0.423), readmission (p=1.00), return to OR within 30 days (p=0.317), return to work (p=0.366), or minor (p=1.00) / major (p=0.480) complication rates. The ID group did demonstrate a statistically-significant lower level of improvement in achieving a 30-percent reduction in leg pain at 3-months vs the DD cohort (p=0.05, RR 0.85, 95% CI 0.7091-0.9996). CONCLUSIONS DD and ID strategies to treat lumbar spondylolisthesis were similar in terms of 12-month PRO improvement, readmission, complications, return to work and satisfaction. The ID cohort did demonstrate a statistically-significant lower reduction in leg pain at 3 months, which was not maintained at 12 months. FDA DEVICE/DRUG STATUS This abstract does not discuss or include any applicable devices or drugs. Debate persists regarding the optimal surgical strategy to treat lumbar spondylolisthesis. The recent development of novel anterior approaches (direct lateral and oblique lumbar interbody fusion) has led to a rise in the utilization of anterior lumbar interbody fusion procedural codes. These new techniques rely on indirect decompression to treat the patient's neural compression, which is criticized by proponents of direct decompressive techniques. To compare 3- and 12-month outcomes between patients with lumbar spondylolisthesis treated with direct decompression (DD) vs indirect decompression (ID) and interbody fusion. Lumbar module of the Quality Outcomes Database (QOD), a national, multicenter prospective spine registry. A total of 4,163 patients underwent DD (posterior lumbar laminectomy with interbody fusion) and 86 patients underwent ID (anterior lumbar interbody fusion and posterior instrumentation/fusion without laminectomy). Oswestry Disability Index (ODI), NRS-back pain, NRS-leg pain, EQ5D, satisfaction, major complication, minor complication, readmission, and return to work at 3- and 12-months postsurgery, return to OR within 30 days. Propensity scores were estimated using logistic regression with surgical procedure as the dependent variable. Independent variables included the baseline covariates that are potentially associated with outcomes. Based on the logit of estimated propensity score, DD patients were one-to-one matched to ID patients using a greedy matching strategy without replacement. Binary outcomes between ID and DD propensity score matched patients were compared using McNemar's test. The estimates of risk difference and relative risk were determined as well as 95% confidence intervals using asymptotic score methods. Continuous outcomes were compared using two sample paired t-test. In the propensity-matched analysis, the DD and ID cohorts did not differ with respect to their 12 month postoperative improvement in ODI (p=0.262), back pain (p=0.715), leg pain (p=0.144), EQ5D (0.384), satisfaction (0.423), readmission (p=1.00), return to OR within 30 days (p=0.317), return to work (p=0.366), or minor (p=1.00) / major (p=0.480) complication rates. The ID group did demonstrate a statistically-significant lower level of improvement in achieving a 30-percent reduction in leg pain at 3-months vs the DD cohort (p=0.05, RR 0.85, 95% CI 0.7091-0.9996). DD and ID strategies to treat lumbar spondylolisthesis were similar in terms of 12-month PRO improvement, readmission, complications, return to work and satisfaction. The ID cohort did demonstrate a statistically-significant lower reduction in leg pain at 3 months, which was not maintained at 12 months.