Comparison of minimally invasive decompression alone versus minimally invasive short-segment fusion in the setting of adult degenerative lumbar scoliosis: a propensity score–matched analysis

Abstract

OBJECTIVE Patients with degenerative lumbar scoliosis (DLS) and neurogenic pain may be candidates for decompression alone or short-segment fusion. In this study, minimally invasive surgery (MIS) decompression (MIS-D) and MIS short-segment fusion (MIS-SF) in patients with DLS were compared in a propensity score–matched analysis. METHODS The propensity score was calculated using 13 variables: sex, age, BMI, Charlson Comorbidity Index, smoking status, leg pain, back pain, grade 1 spondylolisthesis, lateral spondylolisthesis, multilevel spondylolisthesis, lumbar Cobb angle, pelvic incidence minus lumbar lordosis, and pelvic tilt in a logistic regression model. One-to-one matching was performed to compare perioperative morbidity and patient-reported outcome measures (PROMs). The minimal clinically important difference (MCID) for patients was calculated based on cutoffs of percentage change from baseline: 42.4% for Oswestry Disability Index (ODI), 25.0% for visual analog scale (VAS) low-back pain, and 55.6% for VAS leg pain. RESULTS A total of 113 patients were included in the propensity score calculation, resulting in 31 matched pairs. Perioperative morbidity was significantly reduced for the MIS-D group, including shorter operative duration (91 vs 204 minutes, p < 0.0001), decreased blood loss (22 vs 116 mL, p = 0.0005), and reduced length of stay (2.6 vs 5.1 days, p = 0.0004). Discharge status (home vs rehabilitation), complications, and reoperation rates were similar. Preoperative PROMs were similar, but after 3 months, improvement was significantly higher for the MIS-SF group in the VAS back pain score (−3.4 vs −1.2, p = 0.044) and Veterans RAND 12-Item Health Survey (VR-12) Mental Component Summary (MCS) score (+10.3 vs +1.9, p = 0.009), and after 1 year the MIS-SF group continued to have significantly greater improvement in the VAS back pain score (−3.9 vs −1.2, p = 0.026), ODI score (−23.1 vs −7.4, p = 0.037), 12-Item Short-Form Health Survey MCS score (+6.5 vs −6.5, p = 0.0374), and VR-12 MCS score (+7.6 vs −5.1, p = 0.047). MCID did not differ significantly between the matched groups for VAS back pain, VAS leg pain, or ODI scores (p = 0.38, 0.055, and 0.072, respectively). CONCLUSIONS Patients with DLS undergoing surgery had similar rates of significant improvement after both MIS-D and MIS-SF. For matched patients, tradeoffs were seen for reduced perioperative morbidity for MIS-D versus greater magnitudes of improvement in back pain, disability, and mental health for patients 1 year after MIS-SF. However, rates of MCID were similar, and the small sample size among the matched patients may be subject to patient outliers, limiting generalizability of these results.