Background context:

Indications for revision lumbar fusion are variable, but include recurrent stenosis (RS), adjacent segment disease (ASD), and pseudarthrosis. The efficacy of revision lumbar fusion has been well established, but their outcomes compared to primary procedures is not well documented.

Purpose:

The purpose of this study was to compares surgical and clinical outcomes between 1) revision and primary lumbar fusion, 2) revision lumbar fusion based on indication (ASD, pseudarthrosis, or RS), and 3) revision lumbar fusion based on whether the index procedure included an isolated decompression or decompression with fusion STUDY DESIGN/SETTING: Retrospective single-institution cohort study PATIENT SAMPLE: 4,671 consecutive lumbar fusions from 2011 to 2021, of which 892 (23.6%) were revision procedures. The indication for revision procedures included 502 (56.3%) for ASD, 153 (17.2%) for pseudarthrosis, and 237 (26.6%) for RS. Of the 892 revision procedures, 694 (77.8%) underwent an index fusion while 198 (22.2%) underwent an index decompression without fusion.

Outcome measures:

Hospital readmissions, all-cause reoperation, need for subsequent revision and patient reported outcome measures (PROMs) at baseline, three-months postoperatively, and one-year postoperatively, including the Mental Health Component score (MCS-12) and Physical Health Component score (PCS-12) of the Short Form 12 survey, the Oswestry Disability Index (ODI), and the Visual Analog Scale (VAS) for Back and Leg pain.

Methods:

Patient demographics, comorbidities, surgical characteristics, and outcomes were collected from electronic medical records. Twenty-eight percent of patients had preoperative and postoperative PROMs. A delta PROM score was calculated for the three-month and one-year postoperative timepoints, which was the change from the preoperative to postoperative value. Univariate comparisons were performed to compare revision fusions to primary fusions. Multivariate logistic regression was performed for all-cause reoperation and subsequent revision surgery, while multivariate linear regression was performed for ∆PROMs at three-months and one-year. Revision procedures were then separately regrouped based on indication for revision fusion and whether they underwent a fusion for their index procedure. Univariate comparisons and multivariate linear regressions for ∆PROMs were then repeated based on the new groupings.

Results:

There was no difference in hospital readmission rate (5.38% vs. 4.60%, p=0.372) or length of stay (4.10 days vs. 3.94 days, p=0.129) between revision and primary lumbar fusion, but revision fusions had a higher rate of all-cause reoperation (16.1% vs. 11.2%, p<0.001) and subsequent revision (13.7% vs. 9.71%, p=0.001), which was confirmed on multivariate logistic regression (Odds Ratio (OR): 1.42, p=0.001 and OR: 1.37, p=0.007, respectively). On multivariate analysis, a revision procedure was an independent risk factor for worse improvement ∆ODI, ∆VAS Back, ∆VAS Leg, and ∆PCS-12 and one-year postoperatively. Regardless of the indication for revision lumbar fusion, patients significantly improved in the three-month and one-year postoperative PCS-12, ODI, VAS Back, and VAS Leg, with the exception of the three-month PCS-12 for pseudarthrosis (p=0.620). Patients undergoing revision for ASD had significantly worse one-year postoperative PCS-12 (32.3 vs. Pseudarthrosis: 35.6 and RS: 37.0, p=0.026), but there were no differences in ∆PROMs. There was no difference in hospital readmission, all-cause reoperation, or subsequent revision based on whether a patient had an index lumbar fusion or isolated decompression. Multivariate linear regression analysis found that a surgical indication of pseudarthrosis was a significant predictor of decreased improvement in three-month ∆VAS Leg (ref: ASD, β=2.26, p=0.036), but having an index fusion did not significantly predict worse improvement in ∆PROMs when compared to isolated decompressions.

Conclusions:

Revision lumbar fusions had a higher rate of reoperation and subsequent revision surgery when compared to primary lumbar fusions, but there were no difference in hospital readmission rates. Patients undergoing revision lumbar fusion experience improvements in all patient reported outcome measures, but their baseline, postoperative, and magnitude of improvement are worse than primary procedures. Regardless of whether the lumbar fusion is a primary or revision procedure, all patients have significant improvements in pain, disability and physical function. Further, the indication for the revision procedure is not correlated with the expected magnitude of improvement in patient reported outcomes. Finally, no differences in baseline, postoperative, and ∆PROMs for revision fusions were identified when stratifying by whether the patient had an index decompression or fusion.