Perioperative Outcomes, Complications, and Costs Associated with Lumbar Spinal Fusion in Older Patients with Spinal Stenosis and Spondylolisthesis

Neurosurg Focus 36 (6):E5, 2014 ©AANS, 2014

Perioperative outcomes, complications, and costs associated with lumbar spinal fusion in older patients with spinal stenosis and spondylolisthesis

Kevin L. Ong, Ph.D., P.E.,1 Joshua D. Auerbach, M.D.,2 Edmund Lau, M.S.,3 Jordana Schmier, M.A.,4 and Jorge A. Ochoa, Ph.D., P.E.3 1Exponent, Inc., Philadelphia, Pennsylvania; 2Department of Orthopaedics, Bronx-Lebanon Hospital Center, Albert Einstein College of Medicine, Bronx, New York; 3Exponent, Inc., Menlo Park, California; and 4Exponent, Inc., Alexandria, Virginia

Object. The purpose of this study was to quantify the perioperative outcomes, complications, and costs associated with posterolateral spinal fusion (PSF) among Medicare enrollees with lumbar spinal stenosis (LSS) and/or spondylolisthesis by using a national Medicare claims database. Methods. A 5% systematic sample of Medicare claims data (2005–2009) was used to identify outcomes in patients who had undergone PSF for a diagnosis of LSS and/or spondylolisthesis. Patients eligible for study inclusion also required a minimum of 2 years of follow-up and a claim history of at least 12 months prior to surgery. Results. A final cohort of 1672 patients was eligible for analysis. Approximately half (50.7%) had LSS only, 10.2% had spondylolisthesis only, and 39.1% had both LSS and spondylolisthesis. The average age was 71.4 years, and the average length of stay was 4.6 days. At 3 months and 1 and 2 years postoperatively, the incidence of spine reoperation was 10.9%, 13.3%, and 16.9%, respectively, whereas readmissions for complications occurred in 11.1%, 17.5%, and 24.9% of cases, respectively. At 2 years postoperatively, 36.2% of patients had either undergone spine reoperation and/or received an epidural injection. The average Medicare payment was $36,230 ± $17,020, $46,840 ± $31,350, and $61,610 ± $46,580 at 3 months, 1 year, and 2 years after surgery, respectively. Conclusions. The data showed that 1 in 6 elderly patients treated with PSF for LSS or spondylolisthesis underwent reoperation on the spine within 2 years of surgery, and nearly 1 in 4 patients was readmitted for a surgery-related complication. These data highlight several potential areas in which improvements may be made in the effective delivery and cost of surgical care for patients with spinal stenosis and spondylolisthesis. (http://thejns.org/doi/abs/10.3171/2014.4.FOCUS1440)

Key Words • spine fusion • stenosis • spondylolisthesis • lumbar spine

pinal fusion procedures have become increasingly the reasons for this trend are unclear. A recent review of common in the treatment of various spinal disor- the period from 2002 to 2007 revealed that in patients ders, such as spinal stenosis and spondylolisthesis, with a primary diagnosis of spinal stenosis, treatment via Sto facilitate spinal decompression and stabilization. Fu- complex fusion procedures increased from 1.3 to 19.9 sion is also used to treat degenerative disc disorder but operations per 100,000 Medicare beneficiaries, whereas remains controversial in the absence of other codiagno- treatment via decompression or simple fusion decreased 16 ses, in the elderly, and at multiple levels (3 or more). A by approximately 10% and 20%, respectively.12 recent study of the Healthcare Cost and Utilization Proj- Complex procedures are generally more invasive and ect Nationwide Inpatient Sample database showed that can involve greater risks of complication, which in turn from 1998 to 2008, the annual number of spinal fusion leads to higher health care costs.6,14 Thus, as these com- discharges in the United States increased from 174,223 plex procedures are used more frequently, it is important to 413,171, which corresponds to 64.5 and 135.5 cases 22 to identify their correct indications. There is still a lack of per 100,000 people, respectively. The therapeutic ap- consensus about the surgical approaches and techniques proach to spinal stenosis and spondylolisthesis has also shifted toward more complicated procedures, although that are most appropriate for various types of patients and disease characteristics;12 therefore, surgeon preference remains a significant factor in the type of treatment a pa- Abbreviations used in this paper: CCI = Charlson Comorbid- tient receives.17 ity Index; CPT = Current Procedural Terminology; ICD-9-CM = International Classification of Diseases, Ninth Revision, Clinical Treatments for spinal stenosis and spondylolisthesis Modification; LSS = lumbar spinal stenosis; PSF = posterolateral are varied, ranging from conservative therapies, such as lumbar spinal fusion. noninvasive physical manipulation and physical therapy,

Neurosurg Focus / Volume 36 / June 2014 1

Unauthenticated | Downloaded 10/01/21 02:24 AM UTC K. L. Ong et al. to invasive surgical procedures, such as spinal fusion. For was the Charlson Comorbidity Index (CCI). Patients who patients with symptomatic degenerative lumbar spon- had undergone multilevel fusion were identified as those dylolisthesis, the use of spinal fusion with instrumenta- with CPT code 22614, whereas those who had undergone tion can offer better outcomes than those obtained with posterior instrumentation were identified with CPT codes surgical decompression alone; however, results can vary 22840, 22842, 22843, or 22844. depending on a patient’s history, age, comorbidities, and The length of stay for the index PSF surgery was current disease-related symptoms.11,21,26 For patients with calculated from the Medicare claims data. Discharge spinal stenosis without spondylolisthesis, spinal fusion status—that is, discharge to home, skilled nursing facil- is often performed after decompression to address con- ity, rehabilitation facility, or home health services—was comitant low-back pain from facet arthrosis or degenera- also examined. Patients were followed up for 2 years to tive disc disease. Given the lack of consensus regarding determine the incidence of postoperative complications, the best treatment algorithm for these disorders, there is a which included spine reoperation, infection, lumbosacral continued need to develop surgical strategies and medical spinal cord injury, lumbosacral nerve root injury, dural devices to improve both intraoperative and postoperative tear and/or puncture, pneumonia, pulmonary embolism, treatment and recovery. Particularly because of the wide myocardial infarction, mechanical complication, hemor- variety of therapeutic options available, it is important rhage and/or hematoma, heterotopic ossification, lumbar to understand the burden of current surgical procedures spine fracture, postlaminectomy syndrome, wound infec- on the health care system in terms of perioperative com- tion, urinary tract infection, nonunion or malunion, re- plications and health care costs. Therefore, our purpose moval of hardware, and readmission for associated com- in conducting this study was to quantify the periopera- plications. These complications were identified based on tive outcomes, complications, and costs associated with the presence of the corresponding ICD-9-CM diagnosis posterolateral lumbar spinal fusion (PSF) among patients codes and ICD-9-CM or CPT procedure codes from the with lumbar spinal stenosis (LSS) and spondylolisthesis claims records (Table 2). The number of epidural steroid by using a national Medicare claims database. injections in the 12 months before and the 2 years after PSF, as well as the incidence of postoperative electrical stimulation use and blood transfusion, was also assessed. Methods Treatment costs from a payer’s perspective, defined in our study as Medicare’s actual reimbursement and/or pay- A 5% systematic sample of the Medicare claims data 7 (2005–2009) was used to identify patients who had un- ment (adjusted to May 2011 dollars ), were also evaluated dergone PSF for LSS and/or spondylolisthesis between for the 24 months of follow-up. Payment by the Medicare 2006 and 2007 and to track their outcomes. This sam- trust fund (that is, cost to payer) for care associated with ple contained claims data for approximately 2.5 million these patients was compiled from the Medicare claims Medicare beneficiaries in the US. From this data set, we records. identified patients by using the Current Procedural Ter- Outcomes and treatment costs within the cohort minology (CPT) code 22612 and LSS and spondylolisthe- were compared using logistic regression, adjusting for pa- sis by using the International Classification of Diseases, tient sex, age, census region (Northeast, Midwest, South, Ninth Revision, Clinical Modification (ICD-9-CM) di- West), comorbidities (CCI score), race, socioeconomic agnosis codes 724.02, 738.4, and 756.12. To verify that status (Medicare buy-in status), levels fused (single, mul- patients did not have 360° fusion, we excluded PSFs that tiple), and primary/principal diagnosis (stenosis, spondy- were coded concurrently with an interbody device (CPT lolisthesis, others). The Medicare buy-in status indicated code 22851) or anterior spine fusion (CPT code 22558). patients whose Medicare premiums and deductibles were To limit the amount of incomplete health and claims subsidized by the state because of their financial status history, patients were excluded from the study if they were and was used as a proxy for the patient’s socioeconomic status. To account for the general health status of each pa- non-US residents or if they were not enrolled in Medicare 10 Parts A (hospital providers) and B (physician providers). tient, the composite CCI (Dartmouth-Manitoba version ), To further limit the confounding effects, we also excluded a metric that predicts the risk of 10-year mortality, was patients who did not have a minimum of 2 years of post- calculated. All disease diagnoses in a 12-month period operative follow-up or a complete claims history in the before the index PSF were compiled from the Medicare full 12 months prior to PSF, as well as any patients who claims. The CCI algorithm utilizes 19 categories of dis- had undergone any spine surgery (for example, lumbar eases based on the diagnosis and surgeries indicated in a spine fusion, lumbar total disc replacement, laminectomy patient. A “weight” with values of 1, 2, 3, or 6 is assigned or laminotomy, or any operation involving interspinous to each category, and the final index is a composite value devices) within 12 months prior to the PSF. Such patients representing the overall degree of comorbidity. For this were identified based on the presence of the correspond- analysis, the CCI values were grouped into previously es- tablished categories:19 0, none; 1–2, low; 3–4, moderate; ing ICD-9-CM or CPT codes from the claims records ≥ (Table 1). Medicare beneficiaries younger than 65 years of and 5, high. Statistical significance was defined at p < age were included in the study to provide an overall per- 0.05. spective of the health care burden for PSF patients, even Results though those younger than 65 years have end-stage renal disease or physical disabilities. Patient age was taken into In 2006 and 2007, approximately 90% of 106,365 consideration to account for these medical conditions, as patients with a diagnosis of stenosis and/or spondylolis-

2 Neurosurg Focus / Volume 36 / June 2014

Unauthenticated | Downloaded 10/01/21 02:24 AM UTC Spine fusion for stenosis and spondylolisthesis

TABLE 1: Procedure codes for identifying prior surgical procedures*

Procedure ICD-9-CM & CPT Codes lumbar spine fusion ICD-9-CM: 81.06, 81.07, 81.08, 81.36, 81.37, 81.38, 81.61 (effective to Q3 2005), 84.51; CPT: 22533, 22534, 22558, 22585, 22612, 22614, 22630, 22632, 22840, 22851, 22849 lumbar total disc replacement ICD-9-CM: 84.64, 84.65, 84.68; CPT: 0091T, 0094T, 0097T, 0163T, 0164T, 0165T, 22857, 22862, 22865 laminectomy or laminotomy ICD-9-CM: 03.02, 03.09, 80.51; CPT: 63005, 63012, 63017, 63030, 63042, 63044, 63047 op involving interspinous device ICD-9-CM: 84.58 (effective Q4 2005 to Q3 2007), 84.59 (effective Q4 2005), 84.80 to 84.85 (effective Q4 2007); CPT: 0171T, C1821, 22841

* Q3, Q4 = third quarter, fourth quarter. thesis did not receive operative care (fusion, decompres- were white (91.7% white, 6.0% black, 2.4% others). The sion, laminectomy, laminotomy, or discectomy). After distribution of patients by CCI scores of 0, 1–2, 3–4, and applying inclusion and exclusion criteria, we obtained a ≥ 5 was 31.2%, 49.0%, 15.4%, and 4.4%, respectively. The final cohort of 1672 PSF patients for this study. Lumbar largest group of patients was from the South (40.2%), fol- spinal stenosis and spondylolisthesis were the primary di- lowed by the Midwest (31.2%), West (14.5%), and North- agnoses for 58.7% and 18.9% of the patients, respectively, east (14.1%). and were secondary diagnoses for the remaining patients. For the overall cohort, the average age was 71.4 ± Of the 1672 PSF patients, 50.7% had LSS only, 10.2% 7.9 years (range 34–91 years), and patients younger than had spondylolisthesis only, and 39.1% had both LSS and 65 years accounted for 11.5% of the study cohort (Fig. spondylolisthesis. Instrumented fusion was performed 1). The average length of stay (LOS) was 4.6 ± 3.2 days in 69.8% of the patients, and multilevel fusion was per- (Fig. 2 upper). Patients were routinely discharged to formed in 55.8%. The patients had received an average home, (42.2%) or discharged to a skilled nursing facil- of 2.3 ± 2.9 epidural injections in the 12 months prior ity (18.0%), home health service (19.4%), or rehabilitation to the PSF. The average number of preoperative epidural facility (17.2%; Fig. 3 upper). Length of stay was sig- injections increased from 1.8 for those under 65 years of nificantly associated with age (p = 0.001), with patients age to 3.2 for those 85 years and older. Two-thirds of the 65 years or older tending to have longer hospital stays patients (66.7%) were female, and the majority of patients (Fig. 2 lower). Discharge status was also significantly as-

TABLE 2: Diagnosis and procedure codes for identifying surgical procedures and complications

Procedure ICD-9-CM & CPT Codes spine reop (lumbar spine fusion, laminectomy/ ICD-9-CM: 81.07, 81.08, 81.37, 81.38, 03.01, 03.02, 03.09, 80.51, 03.53, 81.65, 81.66, 81.06, 81.36, decompression, spine fracture repair, 84.51, 86.22, 86.28, 84.64, 84.65, 84.68, 84.58 (effective Q4 2005 to Q3 2007), 84.59 (effective removal of hardware, insertion of interbody Q4 2005), 84.80 to 84.85 (effective Q4 2007); CPT: 22533, 22534, 22612, 22614, 22630, 22632, device, incision & drainage, lumbar total 22840, 22849, 63005, 63012, 63017, 63030, 63042, 63044, 63047, 22305, 22310, 22315, 22325, disc replacement, op involving interspinous 22521, 22524, 22850, 22852, 22855, 22558, 22585, 22851, 10180, 12020, 12021, 13160, 22015, devices) 0091T, 0094T, 0097T, 0163T, 0164T, 0165T, 22857, 22862, 22865, 0171T, C1821, 22841 infection ICD-9-CM: 998.59, 996.67; CPT: 22015, 10180 lumbosacral spinal cord injury (cauda equina) ICD-9-CM: 806.4, 806.6, 952.2, 952.3, 952.4 lumbosacral nerve root injury ICD-9-CM: 953.2, 953.3, 953.5 dural tear/puncture/laceration ICD-9-CM: 349.0, 349.1, 349.2, 349.3, 998.2, V45.2 pneumonia ICD-9-CM: 480-486, 487.0 pulmonary embolism ICD-9-CM: 415.1, 415.11, 415.19 myocardial infarction ICD-9-CM: 410 mechanical complication ICD-9-CM: 996.40, 996.41, 996.43, 996.44, 996.46, 996.47, 996.49 hemorrhage/hematoma/seroma ICD-9-CM: 998.1 heterotopic ossification ICD-9-CM: 728.13 closed lumbar spine fracture ICD-9-CM: 805.4 postlaminectomy syndrome ICD-9-CM: 722.83 surgical wound infection or dehiscence ICD-9-CM: 998.5, 998.3 urinary tract infection ICD-9-CM: 595.0, 595.2, 595.3, 595.8, 595.89, 595.9, 599.0 nonunion/malunion ICD-9-CM: 733.8, 733.81, 733.82 removal of hardware CPT: 22850, 22852, 22855 epidural steroid injections CPT: 62282, 62284, 64483, 64484, 64475, 64622, 64623, 62311, 62319

Neurosurg Focus / Volume 36 / June 2014 3

Unauthenticated | Downloaded 10/01/21 02:24 AM UTC K. L. Ong et al.

Fig. 1. Age distribution (in years) among 1672 patients who underwent PSF. sociated with age (p < 0.001), with 59.9% of patients 65 years or older having nonroutine discharges, as compared with 41.1% of those younger than 65 years (Fig. 3 lower). Length of stay and discharge status were not significantly associated with primary diagnosis. At the 3-month, 1-year, and 2-year follow-ups, the incidences of spine reoperation were 10.9%, 13.3%, and 16.9%, respectively, while the combined incidences of spine reoperation and/or epidural injections were 18.2%, 27.0%, and 36.2%, respectively (Table 3). The incidence of readmission for any complication was 11.1% at 3 Fig. 2. Length of stay distribution among 1672 patients who under- months, 17.5% at 1 year, and 24.9% at 2 years. Spine re- went PSF (upper), categorized by age (lower). The x-axes represent operation and/or epidural injections (p = 0.004) and read- the number of days. y.o. = years old. mission for complications (p = 0.011) were significantly associated with age at the 2-year follow-up, tending to be Given the more invasive nature of fusion surgery, pa- lower for those with an age ≥ 65 years (35.2% vs 43.8% tients who undergo such procedures are more likely to re- for reoperation and/or epidural injections, 24.7% vs 27.1% quire added postsurgical care. Approximately half of the for readmission). patients in our study were discharged in 3–4 days (49.1% The overall average payment for all covered health total). However, after further analyzing the patients by care for patients who underwent PSF was $36,230 ± ± ± age, we found that older patients were more likely to re- $17,020, $46,840 $31,350, and $61,610 $46,580, require longer LOSs than the younger patients (p = 0.001, spectively, at 3 months, 1 year, and 2 years after surgery age < 65 vs ≥ 65 years). In addition, younger patients were and corresponded to an overall cost to Medicare of $60.6 also more likely to be routinely discharged, whereas older million, $78.3 million, and $103.0 million for treating ones were more likely to be discharged to skilled nursing these 1672 patients. and/or rehabilitation facilities. Taken together, these data suggest that older patients require more extensive health Discussion care and therefore incur greater costs to help them re- cover from PSF. Analyzing the 5% Medicare data, we found that only In addition to the surgical recovery time, periopera- a small proportion of patients with LSS and spondylolis- tive and postoperative complications can result from sur- thesis underwent fusion surgery. The relatively low rates gical treatment. In our analysis, we found that at the 2-year of surgical treatment were also somewhat surprising, follow-up, urinary tract infection, spine reoperation and/ especially given that previous clinical trials have gener- or epidural injections, epidural injections, postlaminec- ally shown improved outcomes when patients are treated tomy syndrome, and readmission occurred in at least 1 of surgically rather than nonsurgically,26 especially in those 4 patients. Note, however, that some of the reoperations with LSS28 and spondylolisthesis.27 Furthermore, authors within the first 3 months may have been planned and/or of another study reported that the rates of complex fusion part of staged procedures and perhaps unrelated to com- procedures had increased more than 10-fold in patients plications. We did attempt to account for this factor by with a primary diagnosis of spinal stenosis in the period limiting reoperations to only those with any correspond- from 2002 to 2007, although they also noted a slight de- ing diagnoses of infection, spinal cord injury, nerve root crease in the corresponding rate of simple fusion.12 injury, dural tear, mechanical complications, hemorrhage/

4 Neurosurg Focus / Volume 36 / June 2014

Unauthenticated | Downloaded 10/01/21 02:24 AM UTC Spine fusion for stenosis and spondylolisthesis

TABLE 3: Complication rate at the 3-month, 1-year, and 2-year follow-ups*

Complication 3-Mo FU 1-Yr FU 2-Yr FU spine reop 10.9% 13.3% 16.9% epidural injection 8.61% 17.1% 26.1% spine reop &/or epidural 18.2% 27.0% 36.2% injection infection 3.83% 4.90% 5.92% spinal cord injury 0.60% 0.66% 0.78% nerve root injury 0.06% 0.12% 0.12% dural tear/puncture 5.02% 5.50% 5.98% pneumonia 4.43% 7.78% 13.1% pulmonary embolism 1.85% 2.51% 3.59% myocardial infarction 1.97% 2.81% 4.49% mechanical complication 3.47% 5.50% 8.49% hemorrhage/hematoma 2.81% 3.89% 4.84% heterotopic ossification NA NA NA lumbar spine fracture 2.45% 3.29% 4.19% postlaminectomy syndrome 11.8% 19.1% 25.4% wound infection 4.61% 6.10% 7.00% urinary tract infection 17.9% 30.3% 40.9% nonunion/malunion 1.73% 3.05% 4.13% readmission (for any of the 11.1% 17.5% 24.9% Fig. 3. Discharge status distribution of 1672 patients who underwent above complications) PSF (upper), categorized by age (lower). HHS = home health service; removal of hardware 1.56% 2.51% 4.25% Oth/Unk = other or unknown; Rehab = rehabilitation facility; SNF = skilled nursing facility. * FU = follow-up; NA = not applicable. hematoma/seroma, spine fracture, postlaminectomy syndrome, wound infection, nonunion/malunion, complica- address the previously treated spine levels or adjacent lev- tions due to internal orthopedic device/implant, cauda els. Epidural injections are an area for further investiga- equina syndrome, acquired kyphosis postlaminectomy, or tion into whether improvements can be made in terms of closed thoracolumbar vertebral dislocation. effective delivery and lowered costs of surgical care, since Other reports on PSF have also discussed some of the some researchers have noted that limited data suggest complications we found. For example, revision surgery that epidural steroid injections are effective for improv- rates of 9.3% and pseudarthrosis rates of 4.65% have been ing selected short-term outcomes in some patients with lumbar spinal stenosis,2 whereas others have questioned documented but only after a follow-up of approximately 5 13 years.1 In another study, Cheng et al.4 reported a relatively whether the benefits outweigh the risks. Our study also higher nonunion rate of 19.7% in a patient cohort 4 years demonstrated a readmission rate of 11.1% at 3 months for after PSF. Hardware removal rates of up to 32% have also the complications listed in Table 3, which is another po- been described.5,9,15 Regarding infection, in a multicenter tential area for improving delivery of care in patients who randomized study comparing 2-year outcomes in patients undergo PSF. As a result of surgical complications and undergoing noninstrumented and instrumented postero- postoperative rehabilitation and treatment, the burden of lateral fusion, the latter patients had a higher infection care for our patient cohort averaged $61,610 ± $46,580 at rate of 7.4%, compared with 1.4% in the former.15 The 2 years after PSF and corresponded to an overall 2-year instrumented patients also had a higher rate of nerve root cost to Medicare of $103.0 million for treating these pa- pain (4.4% vs 0%). No instances of minor dural tears were tients, which extrapolates to an estimated $2 billion when found in either cohort. Moreover, relatively low dural tear considering the entire Medicare population. rates of 1.6% were reported in a cohort of PSF patients Lumbar spine fusion has been considered the stan- with severe, chronic low-back pain from spondylolisthe- dard of care for degenerative spondylolisthesis, but fusion sis Grades 1 and 2 or from primary or secondary degen- is also commonly performed in patients with predomi- erative segmental instability.25 In contrast, in our study nant low-back pain and spinal stenosis without spondylo- we found a higher rate of dural tears of approximately 6% listhesis.8,16 More recently, clinicians have suggested that at 2 years post-PSF. motion-preserving interlaminar stabilization could be an It is notable that a relatively large number of the PSF effective treatment for patients with degenerative spondy- patients in our study continued to receive epidural injec- lolisthesis or stenosis.8,20,23 As part of a prospective, ran- tions, but it is unclear if these injections were intended to domized, multicenter clinical trial comparing outcomes

Neurosurg Focus / Volume 36 / June 2014 5

Unauthenticated | Downloaded 10/01/21 02:24 AM UTC K. L. Ong et al. of PSF and interlaminar stabilization, a subset of 150 care database presents a “big picture” view of complica- patients with Grade 1 spondylolisthesis were followed up tions associated with PSF and is an unmatched resource, for a minimum of 2 years.8 Interlaminar stabilization was despite the inability to explore patient-reported or func- found to be safe and as efficacious as PSF with similar tional outcomes. clinical outcomes, while requiring fewer health care re- sources because of reduced LOSs, operative times, and Conclusions blood loss. The PSF patients were also noted to have significantly increased superior-and inferior-level angulation Our analysis of a subset of Medicare patients treated and translation, as compared with the interlaminar stabi- for LSS and/or spondylolisthesis via PSF suggested that lization patients who experienced no significant adjacent- older patients may require more extensive postoperative or index-level radiographic changes. These changes in the care and may be more predisposed to certain types of PSF group could lead to higher risks of adjacent-segment surgical complications. As the rate of surgery increases degeneration. In a case series of 22 patients who had un- and as surgical procedures and medical technologies dergone a combination of unilateral microdecompression become more complicated, medical costs will continue and interspinous distraction device insertion for spinal ste- to increase and place a significant burden on Medicare. nosis, significant improvements in the visual analog scale Therefore, it is important to continue studying outcomes score and Oswestry Disability Index were observed.20 No to better understand which subsets of patients respond major complications were encountered in this cohort. In most favorably to specific types of surgery to facilitate a 2-year study of patients with spinal stenosis and neuro- recovery and minimize costs to both the patient and the genic claudication, patients were randomized to decom- health care system. Our data also highlighted several po- pressive surgery or implantation of the X-STOP interspi- tential areas in which improvements can be made in the nous spacer. 24 Although similar results were achieved in effective delivery and lowered cost of surgical care for both groups, the investigators reported a higher number patients with spinal stenosis and spondylolisthesis. of reoperations in the X-STOP group. In another study Acknowledgment with a follow-up to 2 years, laminectomy was a more cost- effective treatment strategy for patients with lumbar ste- We thank Jeffrey Zigler, J.D., for contributing to the study nosis than X-STOP, with conservative treatment being the design and data review. 3 least cost-effective. At this time, studies of interlaminar Disclosure stabilization have relatively short follow-ups, and thus it is unknown whether the improvement in outcomes will be Dr. Auerbach is a consultant for Paradigm Spine, Medacta International, Zyga Technology Inc., and Medical Metrics Inc. and demonstrated with longer-term data. holds a patent with Medacta International. Exponent, Inc. received Our analysis of the Medicare PSF patient population research funding from Paradigm Spine for this study. The funding had several limitations. Although we were able to control source was not involved in the design and conduct of the study; for comorbid conditions and exclude patients who had un- collection, management, analysis, and interpretation of the data; dergone a previous surgery, we were unable to analyze or preparation and approval of the manuscript. The funding source either the specific types of instrumentation used in PSF reviewed the manuscript. The authors report no conflict of interest or the number of levels treated. In addition, we were un- concerning the materials or methods used in this study or the find- able to ascertain the accuracy or reliability of the cod- ings specified in this paper. Author contributions to the study and manuscript preparation ing of various procedures and diagnoses since we used include the following. Conception and design: Ong, Auerbach, an administrative claims database, although the Medicare Schmier, Ochoa. Acquisition of data: Lau. Analysis and interpreta- 5% data set still represents an important and widely used tion of data: all authors. Drafting the article: Ong. Critically revising resource for research. We were also unable to evaluate the article: Auerbach, Schmier, Ochoa. Reviewed submitted version functional outcomes, such as the Oswestry Disability of manuscript: Ong, Auerbach, Lau. Approved the final version of Index and pain index, because we used the administra- the manuscript on behalf of all authors: Ong. Statistical analysis: tive claims data. Although lumbar spine fusion, which is Lau. Administrative/technical/material support: Ong, Lau, Schmier, Ochoa. Study supervision: Ong, Ochoa. considered the standard of care for degenerative spondylolisthesis, was the focus of our study, patients can also References undergo other surgeries, such as decompression and/or laminectomy. It is unclear to what extent the costs and 1. Barbanti Bròdano G, Lolli F, Martikos K, Gasbarrini A, outcomes may differ between our PSF cohort and a case- Bandiera S, Greggi T, et al: Fueling the debate: Are outcomes better after posterior lumbar interbody fusion (PLIF) or after controlled decompression/laminectomy cohort, although 18 posterolateral fusion (PLF) in adult patients with low-grade previous research has suggested that patients with lum- adult isthmic spondylolisthesis? Evid Based Spine Care J bar spondylolisthesis who undergo decompressive lami- 1:29–34, 2010 nectomy and spinal arthrodesis have lower reoperation 2. Bresnahan BW, Rundell SD, Dagadakis MC, Sullivan SD, Jar- rates but higher overall costs than patients treated with vik JG, Nguyen H, et al: A systematic review to assess com- laminectomy alone. Further research may provide ad- parative effectiveness studies in epidural steroid injections ditional insights into the costs and outcomes for various for lumbar spinal stenosis and to estimate reimbursement amounts. PM R 5:705–714, 2013 treatment modalities. Our cost analysis was limited to 3. Burnett MG, Stein SC, Bartels RH: Cost-effectiveness of cur- direct costs and did not consider indirect costs given the rent treatment strategies for lumbar spinal stenosis: nonsurgi- lack of such data in the administrative claims database. cal care, laminectomy, and X-STOP. Clinical article. J Neu- Despite these limitations, the national scope of the Medi- rosurg Spine 13:39–46, 2010

6 Neurosurg Focus / Volume 36 / June 2014

Unauthenticated | Downloaded 10/01/21 02:24 AM UTC Spine fusion for stenosis and spondylolisthesis

4. Cheng L, Nie L, Zhang L: Posterior lumbar interbody fusion son Index is associated with one-year mortality in emergency versus posterolateral fusion in spondylolisthesis: a prospec- department patients with suspected infection. Acad Emerg tive controlled study in the Han nationality. Int Orthop 33: Med 13:530–536, 2006 1043–1047, 2009 20. Ploumis A, Christodoulou P, Kapoutsis D, Gelalis I, Vraggalas 5. Christensen FB, Thomsen K, Eiskjaer SP, Gelinick J, Bünger V, Beris A: Surgical treatment of lumbar spinal stenosis with CE: Functional outcome after posterolateral spinal fusion us- microdecompression and interspinous distraction device in- ing pedicle screws: comparison between primary and salvage sertion. A case series. J Orthop Surg 7:35, 2012 procedure. Eur Spine J 7:321–327, 1998 21. Radcliff K, Curry P, Hilibrand A, Kepler C, Lurie J, Zhao W, 6. Ciol MA, Deyo RA, Howell E, Kreif S: An assessment of sur- et al: Risk for adjacent segment and same segment reopera- gery for spinal stenosis: time trends, geographic variations, tion after surgery for lumbar stenosis: a subgroup analysis of complications, and reoperations. J Am Geriatr Soc 44:285– the Spine Patient Outcomes Research Trial (SPORT). Spine 290, 1996 (Phila Pa 1976) 38:531–539, 2013 7. Crawford M, Church J (eds): CPI Detailed Report: Data for 22. Rajaee SS, Bae HW, Kanim LE, Delamarter RB: Spinal fu- May 2011. Bureau of Labor Statistics. (http://www.bls.gov/ sion in the United States: analysis of trends from 1998 to cpi/cpid1105.pdf) [Accessed April 21, 2014] 2008. Spine (Phila Pa 1976) 37:67–76, 2012 8. Davis R, Auerbach JD, Bae H, Errico TJ: Can low-grade 23. Sobottke R, Schlüter-Brust K, Kaulhausen T, Röllinghoff M, spondylolisthesis be effectively treated by either coflex in- Joswig B, Stützer H, et al: Interspinous implants (X Stop, terlaminar stabilization or laminectomy and posterior spi- Wallis, Diam) for the treatment of LSS: is there a correlation nal fusion? Two-year clinical and radiographic results from between radiological parameters and clinical outcome? Eur the randomized, prospective, multicenter US investigational Spine J 18:1494–1503, 2009 device exemption trial. Clinical article. J Neurosurg Spine 24. Strömqvist BH, Berg S, Gerdhem P, Johnsson R, Möller A, 19:174–184, 2013 Sahlstrand T, et al: X-stop versus decompressive surgery for 9. Dehoux E, Fourati E, Madi K, Reddy B, Segal P: Postero- lumbar neurogenic intermittent claudication: randomized lateral versus interbody fusion in isthmic spondylolisthesis: controlled trial with 2-year follow-up. Spine (Phila Pa 1976) functional results in 52 cases with a minimum follow-up of 6 38:1436–1442, 2013 years. Acta Orthop Belg 70:578–582, 2004 25. Thomsen K, Christensen FB, Eiskjaer SP, Hansen ES, Fru- 10. Deyo RA, Cherkin DC, Ciol MA: Adapting a clinical comor- ensgaard S, Bünger CE: 1997 Volvo Award Winner in Clini- bidity index for use with ICD-9-CM administrative databases. cal Studies. The effect of pedicle screw instrumentation on J Clin Epidemiol 45:613–619, 1992 functional outcome and fusion rates in posterolateral lumbar 11. Deyo RA, Hickam D, Duckart JP, Piedra M: Complications af- spinal fusion: a prospective, randomized clinical study. Spine ter surgery for lumbar stenosis in a veteran population. Spine (Phila Pa 1976) 22:2813–2822, 1997 (Phila Pa 1976) 38:1695–1702, 2013 26. Watters WC III, Bono CM, Gilbert TJ, Kreiner DS, Mazanec 12. Deyo RA, Mirza SK, Martin BI, Kreuter W, Goodman DC, DJ, Shaffer WO, et al: An evidence-based clinical guideline Jarvik JG: Trends, major medical complications, and charges for the diagnosis and treatment of degenerative lumbar spon- associated with surgery for lumbar spinal stenosis in older dylolisthesis. Spine J 9:609–614, 2009 adults. JAMA 303:1259–1265, 2010 27. Weinstein JN, Lurie JD, Tosteson TD, Zhao W, Blood EA, 13. Epstein NE: The risks of epidural and transforaminal steroid Tosteson AN, et al: Surgical compared with nonoperative injections in the Spine: commentary and a comprehensive re- treatment for lumbar degenerative spondylolisthesis. Four- view of the literature. Surg Neurol Int 4 (Suppl 2):S74–S93, year results in the Spine Patient Outcomes Research Trial 2013 (SPORT) randomized and observational cohorts. J Bone 14. Fritzell P, Hägg O, Nordwall A: Complications in lumbar fu- Joint Surg Am 91:1295–1304, 2009 sion surgery for chronic low back pain: comparison of three 28. Weinstein JN, Tosteson TD, Lurie JD, Tosteson A, Blood E, surgical techniques used in a prospective randomized study. Herkowitz H, et al: Surgical versus nonoperative treatment for A report from the Swedish Lumbar Spine Study Group. Eur lumbar spinal stenosis four-year results of the Spine Patient Spine J 12:178–189, 2003 Outcomes Research Trial. Spine (Phila Pa 1976) 35:1329– 15. Fritzell P, Hägg O, Wessberg P, Nordwall A: 2001 Volvo 1338, 2010 Award Winner in Clinical Studies. Lumbar fusion versus non- surgical treatment for chronic low back pain: a multicenter randomized controlled trial from the Swedish Lumbar Spine Study Group. Spine (Phila Pa 1976) 26:2521–2534, 2001 Manuscript submitted February 15, 2014. 16. International Society for the Advancement of Spine Sur- Accepted April 3, 2014. gery (ISASS): Policy statement on lumbar spinal fusion Portions of this analysis have been previously presented at the surgery. ISASS.org. (https://www.isass.org/public_policy/ International Society of Pharmacoeconomics and Outcomes Re- 2011-07-15_policy_statement_lumbar_surgery.html) [Ac- search 17th Annual International Meeting held in Washington, DC, cessed April 21, 2014] on June 2–6, 2012; the North American Spine Society 27th Annual 17. Katz JN, Lipson SJ, Lew RA, Grobler LJ, Weinstein JN, Brick Meeting held in Dallas, Texas, on October 24–27, 2012; SpineWeek GW, et al: Lumbar laminectomy alone or with instrumented 2012 held in Amsterdam, Netherlands, on May 28–June 1, 2012; or noninstrumented arthrodesis in degenerative lumbar spi- and the International Society for the Advancement of Spine Surgery nal stenosis. Patient selection, costs, and surgical outcomes. 12th Annual Conference held in Barcelona, Spain, on March 20–23, Spine (Phila Pa 1976) 22:1123–1131, 1997 2012. 18. Lad SP, Babu R, Baker AA, Ugiliweneza B, Kong M, Bagley Please include this information when citing this paper: DOI: CA, et al: Complications, reoperation rates, and health-care 10.3171/2014.4.FOCUS1440. cost following surgical treatment of lumbar spondylolisthesis. Address correspondence to: Kevin Ong, Ph.D., P.E., Exponent J Bone Joint Surg Am 95:e162, 2013 Inc., 3440 Market St., Ste. 600, Philadelphia, PA 19104. email: 19. Murray SB, Bates DW, Ngo L, Ufberg JW, Shapiro NI: Charl- [email protected].

Neurosurg Focus / Volume 36 / June 2014 7

Unauthenticated | Downloaded 10/01/21 02:24 AM UTC