National Trends in Lumbar Degenerative Spondylolisthesis With Stenosis Treated With Fusion Versus Decompression
Article information
Abstract
Objective
The purpose of this study is to describe utilization, demographics, complications, and revisions for patients with degenerative spondylolisthesis (DS) with stenosis undergoing decompression or decompression with fusion in the United States.
Methods
A national insurance database was used to identify patients who underwent either decompression and fusion or decompression alone for management of DS from 2010–2022. Utilization trends, demographics, and complications for each procedure were compared.
Results
A total of 162,878 patients were identified, of which 78,043 patients underwent combined single-level lumbar decompression and fusion and 84,835 underwent single-level lumbar decompression alone. Between 2010–2021, lumbar decompression and fusion became the predominant surgical intervention for DS in 2016 and continued to account for more than half of all procedures during the remainder of the study period. Factors such as age, sex, comorbidities, geographic region, and physician specialty training were associated with procedure choice. Decompression with fusion was associated with a lower risk of revision surgery up to 5 years postoperatively and an overall lower incidence of 30-day complications.
Conclusion
Decompression with fusion has become the most common treatment for lumbar DS over the past decade despite a lack of compelling evidence supporting its use compared to decompression alone. A variety of patient and surgeon-specific factors is associated with procedure choice. After accounting for cofounders, we identified treatment-specific complications that may be valuable when counseling patients.
INTRODUCTION
Degenerative spondylolisthesis (DS) with stenosis is the forward slippage of one vertebra relative to the adjacent segment with associated central canal narrowing often secondary to facet hypertrophy, ligamentum flavum hypertrophy, and disc bulging [1]. Patients typically present after 50 years of age with primary complaints of low back pain and radicular leg pain [2]. Although the prevalence of DS is variable by gender and ethnicity, it has been reported to be as high as 25% in elderly patients [3]. Natural history studies suggest that the majority of patients with DS can be managed nonoperatively, especially among patients without symptoms upon presentation [4]. However, the SPORT (Spine Patient Outcomes Research Trial) demonstrated that for patients who fail initial conservative treatment, surgical intervention can result in significantly improved outcomes at 2 years postoperatively [5].
Despite the benefits of surgical management, there remains significant controversy whether decompression alone or decompression with fusion is the optimal treatment strategy. Literature from the 1990s and early 2000s suggested that decompression with fusion was superior. In a prospective randomized study by Herkowitz and Kurz [6], patients undergoing decompression with fusion had significantly less slip progression with improved patient reported outcomes at 3 years even among those with pseudoarthrosis. Additionally, Kornblum et al. [7] reported that at a mean of 7 years postoperatively, 86% of patients who underwent decompression with fusion reported excellent to good results compared to only 56% of patients with decompression alone. Lastly, Bridwell et al. [8] found that in a prospectively enrolled cohort of 44 patients, nearly 50% treated with decompression alone had progression of disease and only 33% had functional improvement at ≥2 years postoperatively.
More recently, there has been renewed interest in performing decompression alone for DS, especially in select patient populations. In a study by Kelleher et al. [9], minimally invasive decompression alone was performed for patients with leg-dominant radicular symptoms without dynamic instability on preoperative radiographs. Patients in their study had significant improvement in Oswestry Disability Index scores with only 4% of patients requiring revision at a mean of 31.8 months. In the randomized trial by Austevoll et al. [10], decompression only procedures were found to be noninferior to decompression with fusion. Furthermore, operative time, hospital length of stay, and complications were all found to be reduced in the decompression alone cohort. As discussed by Joaquim et al. [11] excluding patients with high-risk phenotypes such as increased facet angle, hypermobility, and predominant low back pain symptoms may allow for more appropriate selection of patients who are ideal candidates for decompression only procedures.
Given the ongoing evolution of treatment for patients with DS, an updated analysis of nationwide utilization and demographic data is warranted. Additionally, analysis of postoperative outcomes and complications can inform surgical decision making and patient counseling. The purpose of this study was to describe utilization, demographics, revisions, and complications for patients with DS undergoing decompression or decompression with fusion in the United States.
MATERIALS AND METHODS
The PearlDiver Mariner Database (PearlDiver Technologies, Fort Wayne, IN, USA) was queried to identify patients from 2010 to 2022 who underwent single-level combined lumbar decompression and fusion or lumbar decompression alone for degenerative lumbar spondylolisthesis. PearlDiver provides over 45 billion HIPAA (Health Insurance Portability and Accountability Act)-compliant inpatient and outpatient records, and relevant demographics, diagnoses, and procedures from across all payer types. International Classification of Diseases 9th and 10th edition and Current Procedural Terminology Codes (Supplementary Table 1) were used to identify relevant demographic and procedural data. Patients associated with any of the following criteria were excluded: age <18 years, history of lumbar spine surgery, rheumatoid arthritis, cauda equina, spinal instability, deformity, isthmic spondylolisthesis, surgical indications for trauma, malignancy, or infection, corpectomy or osteotomy procedures. Table 1 displays the demographic and procedure characteristics for the patients who underwent combined lumbar decompression and fusion or lumbar decompression alone. Patients with obesity were identified and classified as having a body mass index (BMI) >30 kg/m2. Continuous variables such as age, and Elixhauser Comorbidity Index (ECI) were recorded as mean values with standard deviation while categorical variables were recorded as percentages. Annual utilization trends from 2010 to 2021 were also recorded as raw values and percentages. Procedure utilization data from 2022 was excluded due to the incompleteness of the dataset at the time of analysis. Informed consent and Institutional Review Board approval were not required as all patient data is anonymized, and research is exempt.
Demographic and procedural characteristics for 1-level lumbar fusion and lumbar decompression in the treatment of lumbar degenerative spondylolisthesis
Nearest-neighbor propensity-score matching was utilized to balance the treatment groups on confounding variables such as age, sex, ECI, BMI, smoking, preoperative opioid usage, osteoporosis and other potential confounding comorbidities listed in Table 1. Opioid utilization was examined at 3, 90, 180, and 365 days postoperatively, and utilization at each timepoint were assessed independently. Complication and readmission rates were assessed within 30 days postoperatively, with complications categorized as surgical or medical. Additionally, revision rates, defined as subsequent single-level lumbar fusion, were recorded at 1, 2, 3, 4, and 5-years postoperatively and analyzed independently.
Group differences in demographics, comorbidities, and procedural characteristics were assessed with Student t-tests and Pearson chi-square tests, where appropriate. After propensityscore matching, the analysis was repeated to confirm successful balancing of covariates between surgical treatment groups.
Conditional logistic regression analysis was performed to evaluate whether lumbar decompression fusion with decompression or lumbar decompression alone was predictive of annual revision rates from 1 to 5 years postoperatively, opioid use at various time points, or 30-day complications and readmission. All statistical analyses were conducted using Rstudio (ver. 4.4.2) within the PearlDiver Mariner Database program. A pvalue <0.05 indicates statistical significance.
RESULTS
From 2010 to 2022, a total of 162,878 patients were identified, of which 78,043 patients underwent combined single-level lumbar decompression and fusion (60.7% female) and 84,835 underwent single-level lumbar decompression alone (54.1% female). On average, patients who underwent combined lumbar decompression and fusion surgery were younger (62.5±10.9 years vs. 64.2±11.3 years, p<0.001), and had fewer comorbidities (ECI 4.6±3.4 vs. 4.8±3.3, p<0.001) than patients who underwent lumbar decompression alone (Table 1). However, a greater percentage of patients in the combined lumbar decompression and fusion group were obese (38.23% vs. 37.58%, p=0.007) (Table 1). After stratifying by obesity status, obese patients were more likely to undergo a decompression only (51.68%) compared to a lumbar decompression and fusion (48.32%) (p<0.001) (Table 2).
Regional analysis of cases associated with obesity, stratified by operative approach for degenerative lumbar spondylolisthesis
Across the 10-year study period, regional utilization trends demonstrated that lumbar decompression alone was the most frequently utilized procedure in all geographic regions. The largest differences in utilization between decompression alone and decompression and fusion occurred in the Midwest (54.2% vs. 45.8%, p<0.001) and Northeast (52.8% vs. 47.2%, p<0.001) but were more similar in the South (50.5% vs. 49.5%, p<0.001) and West (51.6% vs. 48.8%, p<0.001). Overall, the South had the highest volume of lumbar decompression as well as combined lumbar decompression and fusion procedures. Compared to combined lumbar fusion and decompression procedures, a larger proportion of lumbar decompression surgeries were performed in an outpatient setting (39.5% vs. 9.9%, p<0.001). In contrast, there was a larger proportion of lumbar fusion and decompression procedures performed in an inpatient setting compared to lumbar decompressions (89.8% vs. 60.2%, p<0.001). Of the surgeons performing these procedures, neurosurgeons were more likely to perform decompression alone compared to orthopedic surgeons (57.4% vs. 48.5%, p<0.001) (Table 1).
1. Annual Utilization Trends
Annual utilization trends for single-level lumbar decompression and fusion and lumbar decompression alone from 2010 to 2021 are outlined in Tables 3 and 4 along with annual percent change in surgical procedure utilization calculated from the previous year. Change in overall volume between 2010–2021 revealed a +387.1% increase in lumbar decompression with fusion compared to only +79.6% increase in lumbar decompression alone. Furthermore, lumbar decompression and fusion became the predominant surgical intervention for DS in 2016 and continued to account for more than half of all procedures performed each year during the remainder of the study period.
Annual utilization trends for lumbar fusion and lumbar decompression cohorts
Annual utilization trends by region (Midwest, Northeast, West, South)
The largest increases in lumbar decompression and fusion case volumes occurred between 2014 and 2015 with a +24.4% increase, and from 2017 to 2018, increasing by +38.3%. Procedure utilization increased steadily between 2010 to 2019, demonstrating an average annual increase of +17.0%. Lumbar decompression and fusion reached the highest case volume by the end of 2019 (n=12,541) before subsequently dropping in 2020 by -18.2% from the preceding year corresponding to the start of the COVID-19 pandemic. During the following year in 2021, there was a slight increase in case volume (n=10,648) by +0.3%.
Lumbar decompression procedures increased between 2010 and 2015, showing an average annual increase of 7.1%. However, procedure utilization decreased by -49.5% in 2016, reaching the second lowest case volume by the end of that year (n=4,759). The largest increase in procedure utilization occurred between 2016 to 2017, increasing by +15.7%, and between 2017 to 2018 which increased by +39.2%. In 2020, case volumes for lumbar decompression procedures alone decreased by 18.4% by the end of the year (n=8,901) and by 1.6% by the end of 2021 (n=8,757). Despite the annual increase in volume between 2010– 2021, there was a relative 23.9% decline in utilization of decompression procedures for eligible cases.
Fig. 1 demonstrates annual utilization trends from 2010 to 2021 for combined lumbar decompression and fusion and lumbar decompression alone. Linear regression demonstrated a significant overall increase in the number of cases for both procedures, with an average increase of 1,045 combined lumbar decompression and fusion cases per year and an average increase of 460 isolated lumbar decompression procedures per year over this period (p<0.001). Similar trends were observed for each region included in the study (Fig. 2).
Annual utilization trends for lumbar fusion and lumbar decompression alone cases from 2010 to 2021.
Annual utilization trends by region (Midwest, Northeast, West, South) with trendlines for combined fusion and decompression procedures from 2010 to 2021.
2. Propensity-Matched Analysis
After propensity-score matching, a cohort of 156,086 patients (78,043 patients per group) was available for evaluation of postoperative outcomes including revision surgery, complications, and opioid use. Relative to lumbar decompression alone, patients undergoing combined lumbar decompression and fusion demonstrated significantly lower odds of revision surgery at 1 year (3.0% vs. 3.6%; odds ratio [OR], 0.83; 95% confidence interval [CI], 0.79–0.88; p<0.001), 2 years (1.3% vs. 2.9%; OR, 0.46; 95% CI, 0.42–0.50; p<0.001), 3 years (1.0% vs. 2.1%; OR, 0.50; 95% CI, 0.42–0.50; p<0.001), 4 years (0.7% vs. 1.5%; OR, 0.46; 95% CI, 0.42–0.51; p<0.001), and 5 years (0.5% vs. 1.1%; OR, 0.46; 95% CI, 0.41–0.51; p<0.001) postoperatively (Table 5).
Aggregate outcomes for propensity-matched cohorts
Additionally, undergoing lumbar decompression and fusion demonstrated a significantly lower risk for all 30-day complications (10.8% vs. 12.5%; OR, 0.84; 95% CI, 0.82–0.87; p<0.001), all 30-day surgical complications (5.4% vs. 6.6%; OR, 0.81; 95% CI, 0.77–0.84; p<0.001), and all 30-day medical complications (6.8% vs. 7.9%; OR, 0.66; 95% CI, 0.64–0.69; p<0.001) (Table 5). However, there were no significant differences in 30-day readmission rates. Opioid use was significantly higher in the decompression and fusion group at 30 days postoperatively but was lower at 90 days postoperatively (p<0.001 at both time points).
The incidence of individual complications at 30 days postoperatively was low in both groups but did reveal multiple betweengroup differences (Table 6). The most common complications in both groups were uncomplicated neurological injury and anemia, which comprised 63.2% of overall surgical complications and 36.5% of overall medical complications within the combined lumbar decompression and fusion cohort, and 59.9% and 31.2% within the lumbar decompression and fusion cohort, respectively. Urinary tract infection (UTI) was the third most prevalent complication, constituting 30.3% and 27.8% of all med-ical complications across both groups. Incidence of other surgical complications including major neurological injury, durotomy, seroma, hematoma, and major SSI were low in both groups (Table 6).
30-Day medical and surgical complications for propensity-matched cohorts
DISCUSSION
Lumbar decompression and lumbar decompression with fusion are both widely utilized for management of DS in the United States [12]. In the 1990s and early 2000s, multiple small cohort studies revealed favorable outcomes for patients treated by decompression with fusion; however, there has been renewed interest in decompression alone in select patient populations with DS [6-8,13]. Recent studies, including 3 major randomized controlled trials (RCTs), have demonstrated comparable treatment outcomes and the potential for tremendous healthcare savings with decompression alone [10,14-17]. Given the recent data showing noninferiority of decompression alone for stable, low-grade DS, the current study presents a national utilization analysis which compares utilization, regional differences, revision rates, and 30-day complications across procedures to assess evolving trends in clinical practice. The results of the present study, which is the largest and most up-to-date review of DS treatment trends in the literature, found that the use of decompression with fusion has markedly increased over the past decade. We also identified a variety of factors associated with procedure choice as well as procedure-specific complications.
The most important finding of our study is that the utilization of decompression with fusion for the treatment of DS has substantially increased during the past decade in the US [6-8,18]. In 2010, 30.9% of patients underwent fusion and in 2021, the last year of complete data, 54.8% of patients underwent fusion representing almost a 100% increase over a decade. Utilization trends have remained consistent since 2017, as the percentage of patients undergoing fusion was 53.5%–54.8%. This finding is significant as 2 RCTs published in 2016 demonstrated long-term noninferiority of decompression alone compared to decompression with fusion for the treatment of lumbar DS [16,17]. In response to these findings, Sastry et al. [19] analyzed patients from the National Inpatient Sample (NIS) database and found that decompression with fusion procedures increased from 67.4% to 90.4% between 2016 and 2019, concluding that the results of these RCTs had not resulted in changes in surgical practice. The findings of the our study extend this work to the year 2021, showing that this utilization trend has continued. However, our study likely provides a more accurate estimate of overall utilization of these procedures in the United States: the Peardiver Mariner Database includes inpatient and outpatient procedures, comprising 10% of fusions and 40% of decompressions in our cohort, whereas the NIS database include inpatients only [20].
Our analysis found that patient characteristics, geographic region, and physician specialty training were all associated with procedure choice during the study period. Unsurprisingly, patient characteristics including older age, obesity, higher overall comorbidity burden, and the presence of various individual comorbidities were all associated with a higher likelihood of receiving a decompression alone. These findings agree with prior reports and are plausibly explained by an effort to reduce operative time and blood loss in sicker patients or those with shorter life expectancies [21]. When stratified by region, we found similar trends reflecting a significant increase in the overall volume and relative utilization of decompression with fusion. Notably, these changes were most pronounced in the Midwest and South. While a detailed regional sub analysis was beyond the scope of the present study, various patient factors, physician-factors (e.g., specialty, years of training), and nonmedical considerations such as insurance status and hospital status (e.g., for-profit) have all been previously shown to influence treatment decisions [19,22-26]. Taken together, our results underscore that there is substantial and evolving clinical equipoise among spine surgeons regarding the optimal treatment for DS, which are influenced by a variety of factors. Recognizing and accounting for this variation may help to develop better evidence and practice guidelines.
In addition to utilization data, our study also provides an updated overview of clinical outcomes and complications associated with decompression alone compared to decompression and fusion. Notably, this analysis was performed with a matched cohort to mitigate the effects of confounders. The use of the Peardiver database precluded an assessment of functional outcomes. However, we were able to assess postoperative opioid use, which may serve as a proxy for postoperative pain. More than 50% of patients in both groups were using opiates at 30 days postoperatively but there was a small but significantly higher percentage of patients in the decompression only group continuing to use opiates at 90 and 180 days postoperatively. Still, at 1 year postoperatively, there was no significant difference in opioid utilization between groups. The best available evidence from RCTs in this patient population is in agreement with our findings, generally showing equivalent improvements in pain scores, as well as other patient reported outcomes, at 2 years postoperatively [10,16,17]. However, it should be noted that the Spinal Laminectomy versus Instrumented Pedicle Screw trial did find sustained improvements in quality of life out to 4 years postoperatively in their decompression and fusion group [17]. These results highlight the need for additional studies with mid- and long-term data to develop optimal treatment guidelines.
After accounting for relevant confounders, revision rates in our study were significantly higher in the decompression only group at all time points. For example, the percent of patients undergoing revision surgery at 2 years was 2.9% in the decompression only group and 1.4% in the decompression and fusion group. The reason for revision cannot be ascertained from the available data but could include slip progression, infection, adjacent segment disease, persistent pain, or hardware failure. Additionally, we can only speculate on the clinical and health-system significance of the observed 1.5% difference in revision between these 2 study groups at 2 years, with similar findings at other time points. It is also important to contextualize our revision data in light of the available literature. Clinical trials have provided mixed findings, with rates from 10% to as high as 34% at 2 years postoperatively, and generally have demonstrated higher revision rates in the decompression only cohorts. The stark difference in magnitude between these studies and the present one can be explained by how revision was defined: in our study, revision was defined as subsequent single-level lumbar fusion, whereas the trials mentioned evaluated all secondary surgeries as “revision.”
Higher rates of complications were expected in the decompression and fusion group due to the need for more extensive dissection, hardware instrumentation, and overall operative time. However, we instead found that the rates of all 30-day complications (12.5% vs. 10.7%) and various individual complications were lower in the fusion and decompression group. It is important to note that while these differences were statistically significant, the magnitude of difference was typically small (<1%). Prior studies have documented conflicting results with respect to complication rates for the 2 procedures, which is challenging to interpret given the heterogeneity of how complications were defined as well as follow-up period [27]. However, recent met-analyses suggest lower overall complication rates with decompression only [27]. Despite the findings of the present study, the potential added risks of instrumentation should not be minimized.
This study has several important limitations. The study relied on retrospective data from a large national database. The data may contain coding errors with regards to diagnosis, procedures performed, and complications. A small percentage of patient records may have missing demographic information; therefore, patient cohorts seen in the results may add up to less than 100%. The study was unable to examine specific indications or radiographic parameters which may have influenced the decision to choose decompression versus decompression with fusion and was also unable to assess functional outcomes, postoperative pain scores, or reasons for revision. Additionally, it cannot be determined if revisions occurred at the same level or an adjacent segment. Furthermore, data on postoperative complications was limited to 30 days postoperatively, which may not capture all complications such as infections and wound complications. Lastly, due to limitations of the integrated statistical software of the database, a discrete number of patients included in the logistic regression analyses was not provided and we are limited in our ability to report follow-up rates. Nevertheless, the current study was able to examine a large nationally representative patient sample and provides useful data with regards to utilization trends, postoperative opioid usage, and complication rates which can be useful to both surgeons and patients undergoing operative intervention for DS. Future studies will aim to determine which patients are preferable candidates for decompression compared to decompression with fusion by assessing both indications and functional outcomes for each procedure.
CONCLUSION
Since 2016, lumbar decompression with fusion has become the more popular surgical management for DS. Patient, regional, and surgeon-specific factors are associated with procedure choice. In this dataset, 1-year and 5-year revision rates and 30-day complications are lower in patients undergoing decompression and fusion as compared to decompression alone. With recent data supporting the noninferiority of decompression alone for management of DS, additional large, prospective studies in the US are needed to determine optimal treatment for this patient population.
Supplementary Materials
Supplementary Table 1 can be found via https://doi.org/10.14245/ns.2448624.312.
International Classification of Diseases 9th and 10th edition (ICD-9 and ICD-10) and Current Procedural Terminology (CPT) codes
Notes
Conflict of Interest
The authors have nothing to disclose.
Funding/Support
This study received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.
Author Contribution
Conceptualization: JCW, RJH, RKA; Formal analysis: KK, WJK; Data curation: NH, MA; Methodology: JRB, AT; Project administration: FH; Writing – original draft: JRB, KK, NH, AT, MA; Writing – review & editing: FH, WJK, MCG, JCW, RJH, RKA.