Characteristics and outcomes of patients undergoing lumbar spine surgery for axial back pain in the Michigan Spine Surgery Improvement Collaborative

Abstract

BACKGROUND CONTEXTThe indications for surgical intervention of axial back pain without leg pain for degenerative lumbar disorders have been limited in the literature, as most study designs allow some degree of leg symptoms in the inclusion criteria. PURPOSETo determine the outcome of surgery (decompression only vs. fusion) for pure axial back pain without leg pain. STUDY DESIGN/SETTINGProspectively collected data in the Michigan Spine Surgery Improvement Collaborative (MSSIC). PATIENT SAMPLEPatients with pure axial back pain without leg pain underwent lumbar spine surgery for primary diagnoses of lumbar disc herniation, lumbar stenosis, and isthmic or degenerative spondylolisthesis ≤ grade II. OUTCOME MEASURESMinimally clinically important difference (MCID) for back pain, Numeric Rating Scale of back pain, Patient-Reported Outcomes Measurement Information System Physical Function (PROMIS-PF), MCID of PROMIS-PF, and patient satisfaction on the North American Spine Surgery Patient Satisfaction Index were collected at 90 days, 1 year, and 2 years after surgery. METHODSLog-Poisson generalized estimating equation models were constructed with patient-reported outcomes as the independent variable, reporting adjusted risk ratios (RRadj). RESULTSOf the 388 patients at 90 days, multi-level versus single level lumbar surgery decreased the likelihood of obtaining a MCID in back pain by 15% (RRadj=0.85, p=.038). For every one-unit increase in preoperative back pain, the likelihood for a favorable outcome increased by 8% (RRadj=1.08, p<.001). Of the 326 patients at 1 year, symptom duration > 1 year decreased the likelihood of a MCID in back pain by 16% (RRadj=0.84, p=.041). The probability of obtaining a MCID in back pain increased by 9% (RRadj=1.09, p<.001) for every 1-unit increase in baseline back pain score and by 14% for fusions versus decompression alone (RRadj=1.14, p=.0362). Of the 283 patients at 2 years, the likelihood of obtaining MCID in back pain decreased by 30% for patients with depression (RRadj=0.70, p<.001) and increased by 8% with every one-unit increase in baseline back pain score (RRadj=1.08, p<.001). CONCLUSIONSOnly the severity of preoperative back pain was associated with improvement in MCID in back pain at all time points, suggesting that surgery should be considered for selected patients with severe axial pain without leg pain. Fusion surgery versus decompression alone was associated with improved patient-reported outcomes at 1 year only, but not at the other time points.