Article Text

other Versions

Original research
Analysis of utilization patterns of vertebroplasty and kyphoplasty in the Medicare population
  1. Laxmaiah Manchikanti1,2,
  2. Vidyasagar Pampati3,
  3. Joshua A Hirsch4
  1. 1Pain Management Center of Paducah, Paducah, Kentucky, USA
  2. 2Department of Anesthesiology and Perioperative Medicine, University of Louisville, Louisville, Kentucky, USA
  3. 3Department of Statistics, Pain Management Center of Paducah, Paducah, Kentucky, USA
  4. 4Department of NeuroInterventional Radiology, Massachusetts General Hospital, Harvard Medical School, Boston, Massachusetts, USA
  1. Correspondence to Dr Laxmaiah Manchikanti, Pain Management Center of Paducah, 2831 Lone Oak Road, Paducah, KY 42003, USA; drlm{at}


Objective To determine the pattern of utilization of vertebral augmentation procedures including vertebroplasty and kyphoplasty of the lumbar and thoracic spine in the Medicare population in the USA.

Methods This analysis was performed using a standard 5% national sample of the Centers for Medicare and Medicaid Services physician outpatient billing claims from 2001 to 2008 and the Physician/Supplier Procedure Summary Master File for 2009 and 2010. Overall characteristics of utilization patterns for vertebral augmentation procedures were evaluated using multiple variables.

Results From 2002 to 2010, vertebroplasty procedures decreased overall by 24.6% with an average annual decrease of 3.5% per 100 000 Medicare population; from 2006 to 2010 the declines were 42.4% overall and 12.9% annually. From 2006 to 2010, kyphoplasty procedures increased overall by 0.8% with an annual average increase of 0.2% per 100 000 Medicare population.

Conclusion Analysis of growth patterns of vertebroplasty and kyphoplasty in the Medicare population from 2002 to 2010 illustrates a lack of growth of kyphoplasty and a decline in vertebroplasty procedures.

Statistics from


Vertebroplasty and kyphoplasty are two vertebral augmentation procedures commonly used to treat vertebral compression fractures (VCFs) after failure of conservative management and therapy. While there is controversy with regard to effectiveness and superiority of vertebroplasty and kyphoplasty compared with conservative management, vertebral augmentation procedures have been reported to increase in the USA.1–8 Vertebroplasty has been performed in the USA with an assigned Current Procedural Terminology (CPT) code on a regular basis since 2001, whereas kyphoplasty has been performed consistently since 2006. Vertebroplasty involves percutaneous injection of polymethylmethacrylate (PMMA) into the fractured vertebral body, thereby strengthening the bone and alleviating painful symptoms.8 Kyphoplasty, in contrast, is a modified version of vertebroplasty, classically involving inflation of a balloon to create a cavity into the bone, with the balloon theoretically providing height restoration, creating a hollow space for the PMMA to enter, and thereby potentially reducing the required pressure of the cement injection. Current CPT terminology broadens the term kyphoplasty to include non-balloon devices that are used to create a cavity before PMMA deposition. Prior to the publication of two blinded randomized controlled trials in the New England Journal of Medicine, both procedures were widely believed to be effective treatments for VCFs with multiple studies demonstrating significant postoperative decrease in pain, decrease in tension and improved surgical alignment.5–8 Practices offering both types of vertebral augmentatation became more prevalent over time. Some series reported combined results with seemingly effective pain relief resulting from the procedures.9

VCFs are caused by osteoporosis, metastatic cancer or trauma. While many cases of osteoporosis-induced VCF are innocuous and respond to basic conservative management, patients can experience a significant decline in pulmonary function, mobility, activities of daily living, mood, pulmonary embolism, atelectasis, deep vein thrombosis, chronic pain and increased medical costs,8 demonstrating that conservative therapy is not itself risk- or problem-free.10 Furthermore, as long-term survival of cancer patients increases, pathological vertebral fractures secondary to osseous tumors (primary and metastatic) is becoming an increasingly important cause of disability in the USA7 and augmentation is increasingly used in this population.9 Thus, VCFs may occur in up to 19% of patients after the age of 50, affect approximately 25% of postmenopausal women, and the prevalence increases to 40% for women aged 80 years and older.8 Even though VCFs are less common in older men, they still pose a significant health risk. Most patients experiencing an osteoporotic VCF remain asymptomatic or minimally symptomatic; however, a large number of these patients do experience significant pain resulting in decreased quality of life and disability.10 The two randomized controlled trials and accompanying literature published since 20091–3 have elicited significant debate about the vertebral augmentation procedures in general and vertebroplasty in particular. Many papers since then, along with systematic reviews, have shown rather conflicting results with modest relief and improvement.4–7 Furthermore, numerous evaluations have been performed assessing utilization characteristics of vertebral augmentation procedures.8 ,11–15

In this study we sought to evaluate the use of vertebral augmentation procedures (both vertebroplasty and kyphoplasty in the lumbar and thoracic spine) in the Medicare population and assessed the recent effect of the randomized trials of vertebroplasty with utilization trends.


The data for this study were taken from the standard 5% national sample of the Centers for Medicare and Medicaid Services (CMS) physician outpatient billing claims from 2001 to 2008 and the Physician/Supplier Procedure Summary Master File for 2009 and 2010.16–19 The dataset (2001–2008) is a sample of those enrolled in the fee-for-service Medicare program based on selecting records with specific numbers in positions eight and nine of the health insurance claim number and is generated by CMS. The CMS 5% sample dataset is therefore unbiased and unpredictable in terms of any patient characteristics, but does allow appropriate tracking of patients over time and across databases. Consequently, CMS makes this 5% sample available to researchers. In addition, a 100% dataset is so large that it is not feasible to use for research purposes. Physician/Supplier Procedure Summary Master File data for 2009 and 2010 provide 100% data but do not provide the detailed information given in the CMS 5% national sample. Institutional Review Board (IRB) approval is not required for this type of analysis.

Overall, Medicare enrolled more than 43 million beneficiaries in 2010 and is the single largest healthcare payer in the USA.18 The Medicare dataset therefore includes a large proportion of procedures performed in the USA, including vertebral augmentation procedures. In addition to patient age, the database included the CPT procedure codes; the International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnosis codes; date of service, provider specialty, provider zip code and allowed charges.

To yield data for the entire beneficiary population of Medicare, results from the 5% sample were multiplied by 20. In addition, rates were calculated based on Medicare beneficiaries for the corresponding year and are reported as per 100 000 Medicare beneficiaries. The data were tabulated based on the place of service (hospital outpatient department, ambulatory surgery center or office).

From January 2001 thoracic and lumbar vertebroplasty and associated imaging procedures were assigned unique CPT-4 billing codes. Codes were assigned for kyphoplasty from January 2006. Claims for primary vertebroplasty codes 22520 and 22521 and additional code 22522 and primary kyphoplasty codes 22523 and 22524 and additional procedure code of 22525 were used. This study design should allow for capture of all vertebroplasty and kyphoplasty cases performed and billed to CMS in the period studied. Using these data, physicians’ reported specialty and place of service were also identified. We grouped physician specialties into six categories: diagnostic/interventional radiology, orthopedic surgery, neurosurgery, interventional pain management/pain medicine/anesthesiology, physiatrists and other (including neurologists, internists, emergency department physicians, physicians identified only as members of multispecialty groups and non-physicians).

In addition, diagnostic codes were used from (ICD-9-CM).

Data synthesis

The data were analyzed using SPSS V.9.0 statistical software, Microsoft Access 2003 and Microsoft Excel (2003). The procedure rates were calculated per 100 000 Medicare beneficiaries. Average annual percentage changes were calculated by (((final/initial) to the power of (1/number of years)) −1).


Population characteristics

The characteristics of Medicare beneficiaries and vertebral augmentation procedures are shown in Table 1. The data presented changed from 2001 to 2008 (or 2010 when data were available). Since vertebroplasty and kyphoplasty services were combined in 2006, combined data were assessed after 2006. Overall, Medicare beneficiaries increased 10% whereas average annual increases were 1.4% from 2006 to 2010. In contrast, patients receiving vertebral augmentation procedures increased by 10.4% overall with an annual average increase of 5.1%; however, vertebral augmentation services per 100 000 population declined 11.2% overall, with an annual average decline of 2.9%.

Table 1

Characteristics of Medicare beneficiaries and vertebral augmentation procedures

Utilization characteristics

Table 2 shows the utilization of vertebroplasty and kyphoplasty services by Medicare beneficiaries. From 2002 to 2010, vertebroplasty procedures decreased overall by 24.6% with an average annual decrease of 3.5% per 100 000 Medicare population; from 2006 to 2010 the declines were 42.4% overall and 12.9% annually. However, kyphoplasty procedures increased overall by 0.8% with an annual average increase of 0.2% per 100 000 population from 2006 to 2010. The changes were significantly higher from 2006 to 2008; however, the decline started in 2009 and was significant in 2010.

Table 2

Utilization of vertebroplasty and kyphoplasty services in Medicare beneficiaries

Specialty characteristics

Most of the procedures (84% of vertebroplasties and 97% of kyphoplasties) were performed in a hospital setting in 2010. Very few procedures were performed in ambulatory surgery center settings and almost 12.6% of the procedures were performed in an office setting in 2010.

Table 3

Utilization of vertebroplasty and kyphoplasty services by specialty illustrating the proportion of selected specialties and the number of procedures

Table 3 shows services provided by various specialties. In 2010 almost 61% of vertebroplasty procedures were performed by diagnostic radiologists followed by all other specialties, with less than 11% for each specialty. In contrast, kyphoplasty was performed by orthopedic surgeons in 37% of patients, by diagnostic radiologists in 26% of patients and by neurosurgeons in 24%. All other specialties performed a very small proportion of the procedures. When combined, diagnostic radiologists ranked first with 37% of procedures followed by orthopedic surgeons (29%) and neurosurgeons (18%).

Procedural characteristics by state

Data on the procedural characteristics per state show significant geographic variation based on 2008 rates. States in the <25th percentile (≤139) were Alaska, Hawaii, Vermont, New Jersey, New York, Oregon, Iowa, New Mexico, Nevada, California, West Virginia and Rhode Island; states in the 25–50th percentile (139–210) were Pennsylvania, Minnesota, Colorado, Maryland, Arizona, Massachusetts, Connecticut, Michigan, Illinois, Wisconsin, Maine, Washington, Delaware, Louisiana and Ohio; states in the 50–75th percentile (210–274) were New Hampshire, Washington DC, Utah, Wyoming, Montana, Virginia, Kentucky, Mississippi, Tennessee, Georgia, Idaho, North Carolina and Texas; and states in the >75th percentile (>274) were Alabama, Florida, Missouri, Oklahoma, Nebraska, Arkansas, Indiana, Kansas, South Carolina, South Dakota and North Dakota.

Diagnostic characteristics

Figure 1 shows the primary diagnosis for vertebral augmentation procedures for 2008.

Figure 1

Distribution of primary diagnoses for 2008.


Vertebral augmentation procedures in Medicare beneficiaries in the USA for the management of fractures of the thoracic and lumbar spine increased until 2008, with a decline starting in 2009 and extending through 2010. For vertebroplasty procedures, the rate of decrease per 100 000 population from 2006 to 2010 was 42% with an annual average decline of 12.9%. For kyphoplasty, while there were no declines, there were only small increases from 2006 to 2010 with an increase of 0.8% per 100 000 population with an annual average increase of 0.2%. Thus, this evaluation does not show any dramatic increases, as shown in the past, and also shows significant decreases specifically for vertebroplasty. Further, this evaluation does not correlate with the dramatic increases seen with other interventional techniques including spinal epidural injections, facet joint injections, spinal cord stimulation, spinal fusion, laminectomy, hip replacement, percutaneous coronary angioplasty and evaluation and management services.16 ,17 ,19 As shown in table 2, there were some initial increases, specifically compared with the initial data from 2001, which will provide skewed results as it was not fully implemented. However, when considering the data from 2002, vertebroplasty showed an overall decline of 24.6% through 2010 and an annual average decline of 3.5%. Similarly, kyphoplasty increased from a rate of 120 per 100 000 Medicare population to 139 in 2007, 141 in 2008, and the decline started in 2009 with 135, reducing to 121 in 2010. This significant change has been attributed to the publication of two randomized trials in 2009.1 ,2 Both procedures were affected even though the randomized trials referred only to vertebroplasty.

Vertebral augmentation procedures are performed in multiple settings although most are performed in a hospital setting; only 12% are performed in an office setting. The number of procedures performed in an ambulatory surgery center setting is minimal. Our data suggest that the majority of vertebroplasty procedures are performed by radiologists (69%) while most kyphoplasty procedures are performed by a combination of orthopedic and neurospinal surgeons (61%). All other specialties including interventional pain physicians constituted 11% for vertebroplasty and 9% for kyphoplasty.

A review of the state data shows that Alaska had the lowest utilization rate with 43 in 2008 compared with 0 in 2001. South Dakota and North Dakota topped the list with 514 and 552, respectively, per 100 000 population. Many states had a rate of ≥300 including Nebraska, Arkansas, Indiana, Kansas and South Carolina. States with <100 procedures were Virginia, Wisconsin and Wyoming.

A number of studies have evaluated the patterns of utilization of vertebral augmentation procedures. A recent evaluation by Goz et al8 of a retrospective analysis of utilization trends of vertebral augmentation in California, New York and Florida from 2004 to 2008 showed a continued increase. They also showed that the use of kyphoplasty significantly outweighs the use of vertebroplasty. They concluded that the reason for the increasing use of kyphoplasty was probably financial incentives, the specialty performing kyphoplasty, and perceived safety of cavity creation and perceived effectiveness of vertebral height restoration. Leake et al20 evaluated the trends of inpatient spine augmentation between 2001 and 2008. They showed that there was a 741% increase in the number of hospital discharges for patients who underwent spine augmentation, but the year-to-year rate of increase has been declining since 2001. This evaluation showed that, from 2004 to 2008, over 50 000 inpatient vertebroplasty procedures and over 152 000 inpatient kyphoplasty procedures were performed compared with vertebroplasty.

The results of the current study show a decrease in the number of vertebroplasty procedures and no significant change in the number of kyphoplasty procedures per 100 000 Medicare beneficiaries, with a decline noted in procedures from 2009 probably reflecting changing practice patterns developed in the light of previous randomized controlled trials.1 ,2 Furthermore, significant geographic variation in both procedures was demonstrated.

The evidence for either vertebroplasty or kyphoplasty, while continuing to emerge, needs to be considered in light of the randomized controlled trials published in the New England Journal of Medicine in 2009. As with other areas involving interventional care, a number of discrepancies with evidence synthesis remain including placebo-controlled evaluations and study designs.21–25 In the era of comparative effectiveness and evidence-based medicine with exploding healthcare costs, proper methodology in studies is crucial. As previously indicated, we believe that appropriate placebo design and the generalizable nature of the studies are important.20–25 Many practitioners continue to believe that augmentation can play an important role in palliating patients with VCFs. Proper application of vertebral augmentation will improve patients’ pain and function, reduce drug use and may return them to an independent status, which we believe represents a great benefit for society. However, inappropriate provision of any type of intervention—specifically one which incurs substantial expenses—will not provide any benefit but will harm the patients and deplete resources, thus reducing access. By the same token, inappropriately performed evaluations that lead to inaccurate conclusions may reduce healthcare expenditures but also will increase patient suffering, reduce function, increase drug use and finally impede access to medical care. The healthcare costs and utilization issue is not simply related to interventions such as vertebral augmentation procedures or other interventional techniques but is related to overall healthcare utilization patterns, including opioid use which is leading to more deaths than traffic accidents and the other illicit drugs combined.26–30 However, compared with other interventional techniques of the spine, surgical interventions, evaluation and management services, vertebral augmentation procedures either declined or remained the same over the years.

This study has a number of limitations, including the lack of inclusion of participants in Medicare Advantage plans which comprised approximately 10% of enrollees. However, we have included all patients aged >65 receiving traditional fee-for-service Medicare and also those aged <65. This inclusion is important because patients aged <65 represent a significant proportion of patients receiving other interventional techniques.16 ,17 In addition, there may be substantial variations in the data with the number of procedures performed during the introduction of CPT codes—namely, 2001 for vertebroplasty and 2006 for kyphoplasty—which may reflect an inappropriate increase in the number of procedures performed over the years. Another limitation is that some variation may be related to coding errors and diagnostic ambiguity. Furthermore, we have not evaluated actual cost data.


These results show decreases through 2010 for vertebral augmentation procedures. These declines may accelerate in the following years of 2011 and thereafter due to economic conditions, healthcare reform and a number of other factors.27 ,28 ,30


The authors wish to thank Sekar Edem for assistance in the literature search, Tom Prigge for manuscript review and Tonie M Hatton and Diane E Neihoff, transcriptionists, for their assistance in preparation of the manuscript.


View Abstract


  • Funding The data from the Centers for Medicare and Medicaid Services (CMS) were purchased for $16 000 by the American Society of Interventional Pain Physicians (ASIPP). There was no other funding. Internal resources were used in preparing this manuscript.

  • Competing interests JH is a consultant for CareFusion and serves on the Steering Committee for the KAVIAR trial (volunteer position) and on the Data and Safety Monitoring Board (DSMB): CEEP trial (volunteer position).

  • Patient consent Obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.