An official website of the United States government
NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Martin BI, Tosteson ANA, Lurie JD, et al. Variation in the Care of Surgical Conditions: Spinal Stenosis: A Dartmouth Atlas of Health Care Series [Internet]. Lebanon (NH): The Dartmouth Institute for Health Policy and Clinical Practice; 2014 Oct 28.
Back pain is a leading cause of morbidity and disability in the United States.1 It is estimated that up to 80% of people will experience low back pain at some point during their lifetimes.2,3 At any given point in time, about 26% of U.S. adults have low back pain, and 14% have neck pain.4,5 Back pain that lasts at least two weeks occurs in 13.8% of the U.S. population annually, accounting for 2-3% of all physician visits, and is the second most common reason for hospitalizations.4,6-8
About 30 million people in the United States receive professional medical care for a spine problem each year.9 The prevalence of back pain under treatment remained stable from 1997 to 2005 and then declined slightly between 2006 and 2008, possibly due to a decrease in the use of elective procedures during the economic recession.9,10 However, while the prevalence of back pain has remained fairly stable over time, the percentage of patients who describe their back pain as “chronic” has increased, from less than 5% in 1992 to more than 10% of all patients with back pain in 2006.11,12
Low back pain is a leading cause of both lost productivity and medical expenditures. In 2004, over $100 billion was spent in the United States on medical care associated with spine problems—approximately the same amount spent treating cancer, diabetes, or arthritis.9 On average, patients with back pain have 73% higher health care expenses than patients without back pain (Figure 3).
National expenditures attributed to back and neck problems, 1997-2011.
Wide regional variation in back surgery rates has been reported previously by the Dartmouth Atlas Project.13 More recently, marked variation in Medicare costs for an average episode of care (a series of health care encounters related to an occurrence of back pain) has been reported across the United States. Much of the variation was explained by the type of operation chosen; for example, admissions involving a spinal fusion operation were more expensive than those involving a decompression alone. Decisions about how to treat patients in the short period of time after they were discharged were also highly variable. Whether patients were discharged to skilled nursing facilities, referred to home health agencies, sent home with instructions for self-care, or used rehabilitation services was an important factor in explaining differences in costs.14
Back pain is a complex problem, but a specific type of back pain, lumbar spinal stenosis, provides a good example that helps to summarize the treatment options, epidemiology, and evidence of effectiveness of treatment. Simply defined, lumbar spinal stenosis is a narrowing of the space within the vertebrae (backbone) where spinal nerves pass. This is caused by an abnormal thickening of the tissues surrounding the spinal cord and vertebral bodies (Figure 4). The hallmark of this condition is neurogenic claudication—pain in the leg that occurs while walking— which is relieved by sitting down or bending forward. Symptoms due to stenosis typically progress slowly. However, unlike pain from a disc herniation, symptoms from stenosis rarely resolve, but typically wax and wane over time.15 Among older adults, spinal stenosis is common, thought to affect about 30% of people age 60 and over.16
Spinal stenosis. Thickening of the tissue surrounding the spinal cord is shown in purple.
Spinal stenosis can be difficult to diagnose definitively, in part because there is no reliable test. The diagnosis requires consideration of clinical symptoms, as well as imaging that shows a narrow spinal canal. However, in some patients, a narrow spinal canal causes no symptoms. When combined with a clinical evaluation, imaging studies, such as magnetic resonance imaging (MRI), may help diagnose spinal stenosis. However, radiographic and clinical definitions of stenosis lack consensus,17-19 and these classifications often correlate poorly with patient reports of pain and disability.20-22 In fact, spinal abnormalities revealed by imaging are surprisingly common even among asymptomatic people; studies have shown that, among patients without symptoms of back pain, 21% had spinal stenosis, 17% had spine joint problems, and 19% had other abnormalities of the bones and tissues of the spinal canal.23,24 Although the narrowing of the spinal canal can be measured with imaging, the degree of constriction predicts poorly which patients will benefit from surgery.
There are multiple treatment options for people with lumbar spinal stenosis, including medication and physical therapy, steroid injections, and surgery. While the use of tests and treatments for spinal stenosis has grown in recent years, this increase does not appear to be caused by higher prevalence of the disease;25 rather, patients are receiving higher intensity care.
Given the many different treatment options for spinal stenosis, many patients would benefit from shared decision-making, a formal process of educating patients about the risks and benefits of treatment options and engaging them in decisions that promote care consistent with their values and preferences. The Spine Center at Dartmouth-Hitchcock Medical Center and Dr. James Weinstein led a National Institutes of Health-funded trial, the Spine Patient Outcomes Research Trial (SPORT), that studied ways to give patients the best information possible about the different treatments for back pain and which decisions were associated with the best outcomes. Similar efforts to help patients make the best, most informed decisions—even in the context of a brief clinic visit—have been undertaken by The Decision Laboratory, led by Dr. Glyn Elwyn. These decision support tools aim to help patients and providers compare alternative treatment options, even with complex conditions such as back pain and spinal stenosis (Figure 5).
Option grid for spinal stenosis.
As outlined in this report, patients with spinal stenosis often suffer from chronic pain and disability. While successfully navigating these health problems is difficult, resources are available at:
The American Academy of Orthopaedic Surgeons: orthoinfo.aaos.org/topic.cfm?topic=A00575
National Institute of Arthritis and Musculoskeletal and Skin Diseases: www.niams.nih.gov/Health_Info/Back_Pain/
Dartmouth-Hitchcock Medical Center: www.dartmouth-hitchcock.org/patients-visitors/health-encyclopedia
Medical management: Medications are commonly used as initial therapy for common spinal problems, including spinal stenosis. Typical medications include non-steroidal anti-inflammatory drugs (NSAIDs), analgesics, COX-2 inhibitors, muscle relaxants, certain anticonvulsants (e.g., Neurontin), and certain anti-depression medications (e.g., Cymbalta). Systematic reviews highlight a moderate short-term benefit of most of these drugs to relieve pain, but, like nearly all medications, they also have risks.26-28
Although the use of opioids (also referred to as narcotics) in the treatment of back pain has been discouraged in clinical guidelines, there was a 423% increase in opioid use among people with back problems from 1997 to 2005.25 The availability of more potent opioid analgesics since 1997 has been accompanied by higher rates of opioid-related complications, leading to emergency department visits, psychiatric conditions, and death.26,29-31 It is unclear why these treatment patterns have occurred, as the wide geographic variation in opioid use does not appear to reflect disease prevalence, injury, or surgical procedures.32 Opioid use for the long-term management of chronic spinal problems is not supported by scientific evidence of safety and effectiveness.33,34 While more than half of “regular” prescription opioid users have back pain,35 a report from the Cochrane collaboration states that “… opioids for long-term management of chronic [low back pain] remains questionable” and guidelines from the American College of Physicians and the American Pain Society have called for reassessing patients who fail to respond to a time-limited course of opioids.26,33,36
Non-operative therapy: The benefits of medical care, chiropractic care, physical therapy, and other non-surgical interventions to treat spinal stenosis have not been demonstrated.16,37,38 There is some evidence that exercise improves leg pain and functioning compared to no treatment, but these measures do not alter the natural progression of stenosis.39 Nevertheless, one-third of the patients in the Spine Patient Outcomes Research Trial (SPORT) study’s unstructured non-operative treatment group reported significant improvements in symptoms at four years.40 This may be as due to “tincture of time” as any treatment offered.
Epidural steroid injections: Wide geographic variation in epidural steroid injections has also been reported in the United States. Their use does not correlate to evidence-based indications of sciatica or radiculopathy;41,42 does not reduce the rate of subsequent surgery;43-48 and does not appear to obviate the need for opioids, surgery, or medical visits among the elderly U.S. population.42,49 One recent large randomized trial found no short-term benefit of epidural steroid injection relative to injection with lidocaine, a short-term anesthetic, in treating patients with lumbar spinal stenosis.50
Surgical interventions have evolved from traditional decompression—where the tissue compressing the spinal nerves is removed—to include both more invasive types of operations such as spinal fusion, where the spine bones are fixed together, and less invasive procedures such as minimally invasive or percutaneous (through the skin) decompression. Over 94,000 inpatient operations for lumbar stenosis were performed in the United States in 2011, with national hospital costs exceeding $2.3 billion (according to unpublished data from the Healthcare Cost and Utilization Project). There are no recent studies documenting the rate of ambulatory or outpatient decompression operations for spinal stenosis in the United States.
Figure 6 shows the rates of initial (incident) inpatient decompression and fusion operations for lumbar spinal stenosis among Medicare fee-for-service beneficiaries age 65 and over. These estimates exclude patients who had other spine problems, such as spinal fractures, or a diagnosis of cancer. The rate of spinal fusion operations for stenosis increased 67%, from 31.6 per 100,000 Medicare beneficiaries in 2001 to 52.7 per 100,000 Medicare beneficiaries in 2011.
Trends in rates of spinal decompression and fusion, 2001 to 2011. The rates shown in the figure represent decompression and fusion procedures for patients with a diagnosis of lumbar spinal stenosis among all Medicare fee-for-service beneficiaries.
Surgical decompression of the spinal canal, such as laminectomy (removing the rear piece of the vertebrae), eliminates pressure on the spinal nerve roots. A variety of surgical techniques is used, with recent advances toward minimally invasive and microscopic techniques.51,52 The rates of inpatient decompression among older adults have declined as fusion operations have increased and as decompression is increasingly performed as an outpatient procedure (Figure 6).
During the period comprising 2001 through 2011, inpatient spinal decompressions for lumbar spinal stenosis were performed at a rate of 80.0 per 100,000 Medicare beneficiaries across the United States. The rate varied more than eight-fold among hospital referral regions, from fewer than 35 procedures per 100,000 in the Bronx, New York (25.3), Miami, Florida (31.8), and South Bend, Indiana (34.8) to more than 180 per 100,000 in Mason City, Iowa (216.7), Tacoma, Washington (200.7), and Bloomington, Illinois (186.6) (Figure 7). Rates of spinal decompression were generally higher in the Pacific Northwest and northern Mountain states than in other parts of the country (Map 2).
Inpatient spinal decompression per 100,000 Medicare beneficiaries (2001-11). Rates are adjusted for age, sex, and race. The average was created based on odd-numbered years from 2001 to 2011.
While decompression procedures are well established for treating stenosis, fusion operations are an increasingly popular, but controversial, treatment option (Figure 6).53 Spinal fusion is intended to eliminate back pain by joining together two or more adjacent vertebrae under the theory that stabilization will reduce symptoms. The procedure frequently involves implanting cages, rods, or other instrumentation to join vertebrae together. The evidence of the effectiveness of lumbar fusion surgery for treating spinal stenosis (in the absence of significant curvature, called scoliosis, or slipping of a vertebra, called spondylolisthesis) has not been fully established.54 A meta-analysis by Turner et al was among the first to find little evidence supporting fusion surgery for spinal stenosis, reporting large variation in satisfactory results.55 On the other hand, a European cohort study found that fusion operations led to better patient-reported outcomes compared to decompression alone, although fusion increased the risk for surgical complications and repeat operations.56
The average rate of inpatient spinal fusion for lumbar spinal stenosis during the period from 2001 to 2011 in the United States was 41.1 per 100,000 Medicare beneficiaries. The fusion rate varied by a factor of more than fourteen across hospital referral regions, from 9.2 procedures per 100,000 in Bangor, Maine to 127.5 per 100,000 in Bradenton, Florida (Figure 8). Regions with relatively low fusion rates included Fresno, California (12.5), Alameda County, California (14.9), and Scranton, Pennsylvania (17.1). More than 80 procedures per 100,000 were performed in Grand Rapids, Michigan (89.9), Mason City, Iowa (89.2), and Tyler, Texas (88.5) (Map 3).
Inpatient spinal fusion per 100,000 Medicare beneficiaries (2001-11). Rates are adjusted for age, sex, and race. The average was created based on odd-numbered years from 2001 to 2011.
Two small randomized trials have demonstrated that, on average, surgical decompression improves patient-reported measures of pain, disability, and quality of life compared to non-operative treatments.57,58 Other comparative effectiveness studies examining the potential benefits and harms of surgery for spinal stenosis have helped to provide further guidance. Surgical patients with spinal stenosis in the Maine Lumbar Spine Study (MLSS) had greater improvements in patient-reported measures of pain and function through a ten-year follow-up compared to non-surgical patients; however, the surgical advantage narrowed over time.59 The SPORT study reported better pain relief and functional recovery with surgical treatment than with non-surgical treatment for patients with spinal stenosis.40 Overall, decompression procedures appear to have moderate efficacy for stenosis, but these results lessen with time,53 and many patients still have significant problems. At a median four years of follow-up in one cohort study, 17% of patients had undergone a repeat operation, and 30% reported severe pain.14
Compared to decompression procedures alone, complex fusion operations (defined as those involving combined surgical approaches or multiple vertebral levels) are associated with greater risks of life-threatening complications, mortality, and increased health care utilization.60 After adjustment for age, comorbidity, previous spine surgery, and other features in a Medicare population, the likelihood of a life-threatening complication with complex fusions compared to decompressions was almost three times higher. Rehospitalizations within 30 days occurred for 7.8% of patients undergoing decompression compared to 13.0% having a complex fusion. Among Medicare patients undergoing any type of spine surgery for lumbar stenosis, with or without spondylolisthesis, the two-year reoperation rate was 17%. In addition, 25% were readmitted to the hospital due to a surgery-related complication.61
Risk of death after inpatient surgical procedure for lumbar spinal stenosis.
Table 3 and Figure 10 show the rates of postoperative surgical complications and repeat spine surgeries among Medicare beneficiaries undergoing an initial inpatient lumbar spine operation for spinal stenosis (without spondylolisthesis or scoliosis). On average, during the five-year period after surgery, 16% of patients required a second operation among those with fusion and 14% among those with decompression, even after adjusting for differences in patient age, sex, comorbidity, and hospitalizations in the previous year. Fusion operations were associated with greater risk of wound problems and life-threatening complications within 30 days, as well as a significantly higher rate of all-cause repeat hospitalizations.
Surgical safety outcomes following inpatient operation for lumbar spinal stenosis among Medicare beneficiaries.
Cumulative incidence of repeat lumbar spine operations following initial inpatient decompression or fusion for lumbar spinal stenosis among Medicare fee-for-service beneficiaries, 2000-11.
Interspinious process devices (such as the X-Stop© device, approved by the FDA in late 2005) have recently emerged as a less invasive alternative to decompression procedures. These devices are inserted between the spinous processes of adjacent vertebrae and spread the vertebrae apart to prevent the nerve canals from pressing on the nerves. While industry-sponsored randomized trials suggest an advantage over non-surgical treatments, there are only a few clinical trials comparing them to decompression.62-64 While these devices can be placed using only local anesthetics, their use is associated with a higher incidence of reoperation. Interspinous distraction procedures appear to have fewer life-threatening complications at the time of the operation, but lead to more subsequent revision operations.65,66 Other minimally invasive decompression techniques are under development, but the evidence necessary to support their use remains limited.
Trends and variation in treatments for lumbar spinal stenosis likely reflect an aging population, a lack of consensus about the best treatment options, and changes in surgical technologies. While decompression may remain the gold standard for patients for whom non-operative treatments have failed, it is increasingly performed on an outpatient basis, using minimally invasive techniques, and incorporating spinal spacers. Additionally, some patients may seek to undergo lumbar fusion operations. Some have viewed fusion as obviating the need for additional treatments. Sadly, this does not bear out in observation of readmissions, complications, and repeat spine surgery rates.
The complex and changing treatment options highlight the need for the development of better tools to help patients to make the best, most informed treatment choices. Prior work by the Dartmouth Atlas Project has shown that the marked regional variation in surgery for back pain reflects the local practice styles of spine surgeons. For the individual patient, there is often not a single “right” treatment choice. Each has the potential to benefit the patient, but benefit is not certain. Each also entails the possibility of harm or the need for further surgery. In ideal settings, patients should be informed about these options and given the opportunity to participate in shared decision-making, allowing their values and preferences to guide them to the best decision for them.
Conceptual model for decision support process.
Procedures such as spinal fusion have become increasingly common in recent years. These operations, unfortunately, can result in complications, some requiring readmission to the hospital. These findings point to significant opportunities to improve safety and effectiveness in treating back pain. Long-term surveillance of safety measures and patient-reported outcomes are rare in spine surgery but are critically important for informing patients and other stakeholders about the value of spinal procedures.
Surgery for back pain, especially for patients with spinal stenosis, has changed dramatically in recent years but continues to vary from one region to the next. While surgical outcomes research has provided information about when and how these changes have occurred over time, patients are still subject to the accident of geography. In one region, patients are more likely to be offered decompression; in another, fusion; and in a third, medical management may be more common. We know very little about how the variation in care patterns has affected patients and their lives. While more needs to be done to improve the treatments for spinal stenosis and back pain, there is a more immediate opportunity to improve care by implementing shared decision-making. A higher quality decision-making process would help patients find the choice best aligned with their values and preferences.
Data sources and cohort formation: We examined the 100% sample of the Medicare Provider Analysis and Review (MedPAR) file for patients undergoing an initial inpatient lumbar spinal fusion or decompression operation for spinal stenosis from 2001 to 2011. Each MedPAR claim is coded with up to ten diagnosis and six procedure codes. We searched all of these codes to identify patients with spinal stenosis undergoing decompression or fusion operations. Population data for estimating population-based rates came from the Medicare Denominator file and was stratified by five-year increments of age, sex, and race (black/non-black). We combined data from 2001 through 2011 to estimate age-, sex-, and race-adjusted trends in the rates of decompression and fusion operations for spinal stenosis per 100,000 Medicare beneficiaries.
We identified inpatient admissions among beneficiaries age 65 and older, excluding those who were on social security disability insurance, had eligibility for end-stage renal disease, or were enrolled in Medicare HMO programs (e.g., Medicare Advantage). Admissions associated with the surgical indication of spinal stenosis were identified using a previously published and validated hierarchical coding algorithm (www.researchgate.net/publication/257631899_SPINEDEF_%28Version_6%29 _Coding_definitions_for_characterizing_spine-related_medical_encounters).All inpatient admissions that involved an initial (incident) thoracolumbar, lumbar, or lumbosacral fusion or decompression operation for spinal stenosis from 2001 through 2011 were included. However, admissions that included codes for refusion, artificial disc replacement, corpectomy, osteotomy, and kyphectomy were excluded. We further excluded admissions that contained codes for non-degenerative lumbar spinal admissions, such as spinal fracture, vertebral dislocation, spinal cord injury, cervical or thoracic conditions, and inflammatory spondylopathy. Finally, we excluded admissions associated with codes for accidents, neoplasm, HIV or immune deficiency, intraspinal abscess, or osteomyelitis.
Surgical complications: Orthopaedic device complications, wound problems, life-threatening medical complications, and repeat surgery were ascertained for each patient. To calculate the rate and difference in surgical risk between fusion and decompression, we performed a logistic regression for each type of complication, including variables for patient age, sex, race, comorbidity, and previous hospitalizations. Similarly, we examined differences in long-term rates of repeat spine operation between decompression and fusion. We used a Cox proportional hazard regression model to examine differences in the time until a first reoperation between patients undergoing initial decompression and fusion operations.
Economic analyses: For the economic analyses presented in this report, we updated our previously published analysis of the Medical Expenditure Panel Survey9 with data through 2012 to estimate the treated prevalence and the economic burden of back and neck problems in the United States. MEPS is a household survey of medical expenditures weighted to represent national estimates. We focused on adults (> 17 years) with self-reported neck and back problems mapped to spine-related codes from the International Classification of Disease (ICD-9-CM). Inflation-adjusted, survey-weighted generalized linear regression models, adjusting for age, sex, and Charlson comorbidity, were used to calculate the incremental difference in health care costs between patients with and without spine problems.
Rates of inpatient lumbar decompression and fusion for lumbar spinal stenosis among hospital referral regions (2001-11).
Except where otherwise noted, this work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. To view a copy of this license, visit https://creativecommons.org/licenses/by-nc-nd/4.0/
Your browsing activity is empty.
Activity recording is turned off.
See more...
