Thoracic and Lumbar Vertebroplasties Performed in US Medicare Enrollees, 2001-2005Thoracic and Lumbar Vertebroplasties Performed in US Medicare Enrollees, 2001-2005 FREE

To the Editor: Percutaneous vertebroplasty involves the vertebral injection of polymethylmethacrylate cement. Although some indication that this procedure is safe and effective for treating osteoporotic compression fractures exists,¹ the US Medicare program promulgated no national coverage policies for this procedure after reviewing the available nonrandomized evidence.² Nevertheless, local Medicare contractors in multiple jurisdictions have covered vertebroplasty for various indications since as least 2001. We examined vertebroplasty-use patterns in Medicare patients for 2001-2005.

Using vertebroplasty-related Current Procedural Terminology, 4th Edition (CPT-4), codes 22520 (primary thoracic vertebroplasty) and 22521 (primary lumbar vertebroplasty), we performed cross-sectional analyses of aggregate 2001-2005 fee-for-service data from the Medicare all-age Part B Extract Summary System,³ which excludes denied claims and claims for Medicare managed care enrollees. Annual primary vertebroplasty rates (which exclude additional vertebral levels also treated) were therefore expressed per 100 000 Part B fee-for-service enrollees.

Part B Extract Summary System data are cross-stratified by the billing physician's reported specialty and by the listed place of service. We grouped physician specialties into 5 categories: diagnostic or interventional radiology, orthopedic surgery, neurosurgery, anesthesiology or pain management, and other (including neurologists, physiatrists, internists, emergency department physicians, physicians identified only as members of multispecialty groups, and nonphysicians). We grouped places of service into 4 categories: inpatient hospital settings, outpatient hospital settings, physicians' offices, and ambulatory surgery centers.

Because we analyzed data on 100% of known cases, inferential statistics were not required. This study received institutional review board approval.

Vertebroplasty rates nearly doubled from 2001 to 2005, increasing by 32.3% from 2001 to 2002 alone (Table). However, 2005 rates were only 5.0% higher than those from 2004.

Most procedures were performed by diagnostic or interventional radiologists (Table). The proportion performed by anesthesiologists or pain management specialists increased from 4% to 5% during 2001-2004 to 7.1% in 2005; the proportions performed by other specialties remained stable or declined.

Although outpatient hospital settings were the most common treatment sites, the proportions of procedures performed in physician offices and ambulatory surgery centers increased markedly in 2004-2005 (Table) with varying mixtures of specialist intervention. For example, among office-based procedures from 2005, 37.2% were performed by radiologists, while anesthesiologists or pain management specialists performed 35.9%, and orthopedists performed 19.7%. Among ambulatory surgery center procedures from 2005, anesthesiologists or pain management specialists performed 50.5%, while radiologists performed 37.2% and orthopedists performed 1.8%.

Most of the observed growth—rates nearly doubled from 2002 to 2005—preceded the US Food and Drug Administration's approval of polymethylmethacrylate cement use for vertebroplasty in December 2004.⁴ Growth may better reflect factors including shifts in clinical opinion, patient demand, Medicare coverage policies, and the availability of vertebroplasty relative to that of other treatment approaches. The overall increase in outpatient vertebroplasty may mirror earlier trends seen in the growth of outpatient lumbar spine surgery.⁵

Limitations of our data included a lack of clinical and demographic detail and the potential for coding errors. However, with the exception of transient shortfalls, Medicare claims data may be generally concordant with other population-based clinical procedure data. For example, cataract-procedure volume concordance with record-based data from the Rochester Epidemiology Project was nearly 96% when excluding a circumscribed data shortfall period.⁶ Our inability to capture denied claims, those for patients with Medicare managed care or Part A coverage alone and for vertebroplasties billed as “unspecified procedures,” makes our vertebroplasty volume data conservative. However, if such cases decreased over time, then we may have overestimated the actual growth of vertebroplasty use. Our data may not apply to Medicare managed care or non-Medicare populations with differing clinical presentations. Finally, available CPT-4 codes did not capture volumes of competing alternative procedures (eg, kyphoplasty).

Nevertheless, the increase in the volume of vertebroplasty procedures seen in our study is noteworthy given the expected contribution of the Medicare population to vertebroplasty volumes. This increase, especially regarding procedures performed in nonhospital settings, has uncertain clinical and resource use implications and argues for close tracking of future vertebroplasty practice patterns and outcomes.