| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Health Policy and Practice |
1 From the Department of Radiology, Thomas Jefferson University Hospital, 111 S 11th St, 3390 Gibbon, Philadelphia, PA 19107 (J.A.C., W.B.M., L.P., M.E.S., D.C.L.); and Research Department, American College of Radiology, Reston, Va (J.H.S.). From the 2000 RSNA scientific assembly. Received August 17, 2001; revision requested October 11; final revision received April 23, 2002; accepted May 16. Address correspondence to J.A.C. (e-mail: john.carrino@mail.tju.edu).
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
MATERIALS AND METHODS: By using the 1993, 1996, 1998, and 1999 Medicare Part B claims database, we studied Current Procedural Terminology revision 4 codes used for percutaneous spinal injection procedures, including cervical and lumbar discography, disk aspiration and/or injection, facet and/or perifacet joint injection, and epidural steroid injection. For each of these procedures, volume, reimbursement, and physician specialty participation (categorized as radiology, anesthesiology, surgery, physiatry, and other specialties) for each year were recorded.
RESULTS: Despite an overall increase in spinal injection procedure volume and reimbursement from 1993 to 1999, nonradiologists performed most of these procedures. Epidural steroid and facet joint injections had the highest volume and reimbursement during this time period and were performed almost exclusively by nonradiologists (predominantly anesthesiologists). Radiologists performed more discography procedures than did other specialists in 1993, but participation decreased each year, while anesthesiologist participation increased; as of 1999, anesthesiologists performed more discography procedures than did radiologists. Although radiologists performed more disk aspiration procedures than did other specialists, procedure volume remained low during the period studied.
CONCLUSION: Spinal injection volume and reimbursement have increased substantially in the Medicare population from 1993 to 1999. During this interval, radiologist participation has decreased. Nonradiologists perform most spinal injection procedures.
© RSNA, 2002
Index terms: Interventional procedures, utilization, 30.1269 • Radiology and radiologists, socioeconomic issues • Spine, interventional procedures, 30.1269 • Spine, intervertebral disks, 30.1269
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Precise needle placement maximizes the effectiveness of spinal injection procedures (6,18,21,25–30). Therefore, a crucial component is image guidance, often used in conjunction with contrast material injection to verify optimal needle position. By using image guidance, complications are minimized, and accurate placement of medication can be verified. Radiologists are trained to guide needle placement precisely with the use of image guidance, but spinal injection procedures are performed by a variety of specialty physicians other than radiologists. The purpose of this study was to evaluate spinal injection procedures for trends in volume, reimbursement, and physician specialty participation.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In 1993, there were 36.3 million Medicare beneficiaries in the United States: 33.6 million in traditional fee-for-service Medicare and 2.7 million in Medicare HMOs. In 1996, there were 38.1 million Medicare beneficiaries: 33.2 million in traditional fee-for-service Medicare and 4.9 million in Medicare HMOs. In 1998, there were 38.5 million Medicare beneficiaries: 31.9 million in traditional fee-for-service Medicare and 6.6 million in Medicare HMOs. In 1999, there were 40.4 million Medicare beneficiaries: 33.5 million in traditional fee-for-service Medicare and 6.9 million in Medicare HMOs. These years were chosen because they would potentially provide a window into recent trends. Because services for Medicare HMO patients are generally capitated and not handled directly by Medicare fiscal intermediaries, their records are not included in these files and thus are not included in this study.
One author (W.B.M.) reviewed the CPT-4 codes for each study year (Table 1) to identify those that were believed to reflect the most common types of spinal injection procedures that radiologists perform, including lumbar discography (CPT-4 code 62290), cervical discography (CPT-4 code 62291), disk aspiration and/or injection (CPT-4 code 62287), epidural steroid injection (nonselective epidural injection or selective nerve root injection; CPT-4 code 62289 for lumbar injection or 62298 for cervical or thoracic injection), and facet and/or perifacet joint injection (CPT-4 code 64442) (31–34). These procedure codes remained unchanged from 1993 to 1999.
|
We used the "surgical" procedure codes for injection rather than the radiology supervision and interpretation codes to track specialty providers who actually performed the procedures. The database does not link the imaging and injection codes, so it is not possible to track how often image guidance is used in conjunction with spinal injection procedures. The modifier code "51" is occasionally added to a procedure code to identify additional levels injected at the same time as the primary level. Therefore, if a provider injects two vertebral levels, both unmodified (for the first level) and modified (for the second level) codes would be submitted for reimbursement. We included both modified and unmodified procedure codes in the analyses.
To apply for Medicare reimbursement, health care providers must identify their subspecialty by using one of 107 specialty codes. Claims filed by multispecialty groups represent approximately 1.5% of all claims. This number was fairly consistent from year to year (1.7% for 1993, 1.7% for 1996, 1.3% for 1998, and 1.5% for 1999). They were excluded from our analysis because the providing physician’s specialty was indeterminate. The remaining codes in the data set were categorized as follows: anesthesiology, surgery, physiatry (physical medicine and rehabilitation), radiology, and all other specialties. The surgery category included both orthopedic and neurologic surgeons.
To analyze trends, we determined the volume of each procedure, the change in volume of these procedures, the frequency of claims made by the specialty group described above, and the change in the percentage of claims for each procedure originating from the various specialties. Because these are complete numbers of volume in the Medicare fee-for-service beneficiary population, statistical tests—which test whether inferences from samples to populations are appropriate—were not conducted. As complete populations, all differences are real differences and are not due to fluctuations in samples.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
|
|
|
|
|
|
|
Lumbar discography results are summarized in Table 3. In 1993, 4,520 procedures were performed. Over the subsequent years there was a steady increase in volume, with 5,055 procedures performed in 1996, 8,605 in 1998, and 11,323 in 1999, for an overall increase of 151% from 1993 to 1999. Total reimbursement increased from $463,338 in 1993 to $1,402,790 in 1999. The trends in provider distribution demonstrated substantial changes. From 1993 to 1999, increased participation was shown by anesthesiology (11.2% to 35.2% for lumbar discography), physiatry (2.5% to 6.8%), and all other providers (2.5% to 4.3%). Decreased participation was shown by surgery (39.0% to 21.1%) and radiology (44.8% to 32.6%). All specialties showed consistent trends upward or downward in volume, except for providers in the "other" category. Radiologists performed more lumbar discography procedures in 1993 than did any other specialty; however, in 1999, anesthesiologists performed more (n = 3,985) than did radiologists (n = 3,693).
Cervical discography results are summarized in Table 4. In 1993, 1,163 procedures were performed. Over subsequent years, volume remained relatively stable, with 1,272 procedures in 1996, 1,355 in 1998, and 1,244 in 1999, for an overall increase of 7% from 1993 to 1999. Total reimbursement increased slightly from $111,139 in 1993 to $163,007 in 1999. Provider distribution during the study period showed no consistent trend, except for a continuing decrease in participation by surgical specialties (21.2% of procedures in 1993, 13.7% in 1996, 13.7% in 1998, and 10.3% in 1999) and an increase in participation by physiatry (0% in 1993, 0.8% in 1996, 3.0% in 1998, and 4.9% in 1999). Anesthesiology showed a slight overall increase in participation from 1993 (34.2%) to 1999 (41.9%), whereas radiology showed a slight overall decrease in participation (36.6% in 1993 to 34.1% in 1999). In 1993, the specialty that provided the largest proportion of cervical discography procedures was radiology (36.6%). In 1996, 1998, and 1999, however, the largest proportion was provided by anesthesiology (52.6%, 43.5%, and 41.9%, respectively).
Disk aspiration and/or injection results are summarized in Table 5. In 1993, 539 such procedures were performed. The volume decreased in 1996 to 353 and increased again in 1998 (n = 370) and 1999 (n = 499). Overall, there was a 7% decrease in volume of this procedure from 1993 to 1999, with an associated decrease in reimbursement ($304,915 in 1993 to $248,428 in 1999). The trends in provider distribution showed a consistent increase in participation by anesthesiology (5.6% of procedures in 1993, 6.8% in 1996, 9.2% in 1998, and 20.2% in 1999) and radiology (14.8% in 1993, 28.0% in 1996, 44.6% in 1998, and 43.3% in 1999). Surgical specialties showed a steady decrease in participation (69.2% in 1993, 58.6% in 1996, 42.4% in 1998, and 30.7% in 1999). Physiatry had minimal participation, ranging from 0% to 1.2%. In 1993, surgical specialists performed most of these procedures (69.2%), whereas in 1999, radiologists performed the largest proportion (43.3%).
Lumbar epidural steroid injection results are summarized in Table 6. In 1993, there were 444,514 procedures in this group. Volume of procedures varied over subsequent years, decreasing to 406,219 in 1996, increasing to 635,887 in 1998, and decreasing to 482,184 in 1999 for an overall increase of 8% from 1993 to 1999. Because of the large number of lumbar epidural steroid injections, reimbursement for this category was the highest of all procedures studied, increasing from $37,356,181 in 1993 to $49,922,829 in 1999. In all years studied, by far the largest proportion of lumbar epidural steroid injections was provided by anesthesiology (81.7% in 1993, 74.5% in 1996, 79.1% in 1998, and 85.1% in 1999). Radiology provided a minority of these procedures, the proportion of which remained relatively stable (1.1% in 1993, 2.2% in 1996, 2.3% in 1998, and 1.5% in 1999). Radiology participation was also surpassed by physiatry (4.7% in 1999) and surgery (4.4% in 1999).
Cervical and thoracic epidural steroid injection results are summarized in Table 7. In 1993, there were 10,105 procedures in this group. The volume increased steadily over subsequent years (30,087 in 1996, 44,314 in 1998, and 48,210 in 1999) for an overall increase of 377% from 1993 to 1999. Reimbursement followed this trend, increasing from $1,092,870 in 1993 to $5,592,645 in 1999. The provider distribution during the study period remained relatively stable, with anesthesiologists performing the vast majority (91.2% in 1993, 92.5% in 1996, 91.4% in 1998, and 89.0% in 1999) of procedures. As with lumbar epidural injections, radiologists performed only a small proportion of cervical epidural injections, ranging from 0.7% to 1.3% (1.1% in 1999). Radiology participation was again surpassed by both physiatry (4.2% in 1999) and surgery (2.1% in 1999).
Facet and/or perifacet joint injection results are summarized in Table 8. In 1993, there were 38,122 procedures in this group. The volume of procedures increased steadily over subsequent years (64,495 in 1996, 91,456 in 1998, and 94,134 in 1999) for an overall increase of 147% from 1993 to 1999. Reimbursement during this period reflected this growth, increasing from $3,159,856 in 1993 to $8,554,579 in 1999. Provider distribution for facet and/or perifacet joint injections showed a steady increase for anesthesiology (40.8% in 1993, 51.0% in 1996, 56.7% in 1998, and 59.2% in 1999) and physiatry (6.5% in 1993, 8.3% in 1996, 8.6% in 1998, and 9.2% in 1999). A steady decrease in participation was shown by surgery (29.0% in 1993, 18.8% in 1996, 10.8% in 1998, and 9.9% in 1999) and radiology (8.2% in 1993, 6.4% in 1996, 5.1% in 1998, and 4.5% in 1999). In all years studied, the largest proportion of facet and/or perifacet joint injections was provided by anesthesiology.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In our study, disk aspiration and/or injection was the only procedure that showed decreased volume (-7%) from 1993 to 1999. All other injections showed increased volume, with particularly large increases in cervical and thoracic epidural injection (+377%), lumbar discography (+151%), and facet and/or perifacet joint injection (+147%). Although lumbar epidural steroid injection showed little increase during this period (+8%), the volume (482,184 in 1999) and reimbursement ($49,922,829 in 1999) were by far the highest of the procedures we examined.
Anesthesiologists performed the most spinal injection procedures, especially epidural steroid injections (cervical or thoracic, 89.0% in 1999; lumbar, 85.1% in 1999). Radiologists performed a minority of these epidural injections (cervical or thoracic, 1.1% in 1999; lumbar, 1.5% in 1999), which were also more commonly performed by physiatrists and surgeons. Radiologists also performed a minority of facet joint injections (4.5% in 1999), a procedure which was also dominated by anesthesiologists (59.2% in 1999) and performed more commonly by physiatrists and surgeons. In 1993, radiologists performed more lumbar discography procedures (44.8%) than did any other specialty, but this figure decreased to 32.6% in 1999, whereas anesthesiologist participation increased from 11.2% to 35.2% during this period. A similar but less dramatic trend was seen for cervical discography; in 1993, radiologists (36.6%) performed more procedures than did anesthesiologists (34.2%), whereas in 1999, radiologists (34.1%) performed fewer procedures than did anesthesiologists (41.9%). Although radiologists performed more disk aspiration and/or injection procedures (43.3%) than did other specialists in 1999, few such procedures were performed (499 in 1999). Surgical specialists, including orthopedic and neurologic surgeons, showed a decrease or no change in participation for all procedures from 1993 to 1999. Physiatrists showed increased participation for all procedures from 1993 to 1999, but the proportion of procedures performed by physiatrists remained lower than 10% as of 1999.
There is growing opinion that precise placement is essential for the safe and effective use of spinal intervention with percutaneous needles (18,21,25–30). This requires image guidance and an operator with knowledge of imaging anatomy and experience with performing image-guided procedures. Nevertheless, to our knowledge, there are no data that document the volume of image-guided versus non-image-guided spinal injections. It is accepted that image guidance is required to target and confirm locations for discography and intraarticular facet joint injections. There is uncertainty in the realm of epidural steroid placement, however, regarding the use (or need) of imaging guidance.
Numerous trials have been conducted, but to our knowledge, there have been no randomized controlled trials in which non-image-guided and image-guided epidural steroid injection were directly compared. Findings of multiple studies (3,15,18,21,23) have demonstrated the effectiveness of epidural steroid injection for the diagnosis and treatment of patients with radicular pain. However, studies performed without fluoroscopic guidance are not unanimous in demonstrating the benefits of this procedure (2,4,5,16,35). The broad range of successful outcomes, ranging from 18% to 90%, may be related in part to inaccurate anatomic deposition of medication.
Findings of two important prior studies confirm the importance of image guidance for epidural steroid injection. Carette and coinvestigators (35) studied 158 patients with sciatica. They found that epidural steroid injection affords short-term improvement in leg pain and sensory deficits in patients with sciatica caused by a herniated nucleus pulposis. The injection offers no substantial functional benefit, however, nor does it reduce the need for surgery when compared with a placebo (isotonic saline epidural injection). More recently, Riew and co-investigators (23) studied 55 patients with lumbar radicular pain (sciatica) and found epidural steroid injection to be more effective than anesthetic only epidural injection in obviating the need for decompression surgery.
Both of these studies (23,35) were prospective, randomized, controlled, double-blinded clinical trials. Among the differences in study design, however, was the type of injection technique. Carette et al (35) used a non-image-guided approach, while Riew et al (23) used a fluorscopically guided transforaminal approach. While there may be other factors influencing the variability between these studies, the major difference was injection technique, and the discrepancy of the results support that there is improved effectiveness of image-guided epidural steroid injection.
There are also some hazards associated with non-image-guided epidural steroid injection technique. The incidence of failure to reach the epidural space because of use of a interlaminar approach without image guidance ranges from 13% to 30%, and without image guidance, the wrong interlaminar space may be targeted by one or more levels (25,29,30,36). Also, there is a 9% incidence of inadvertent epidural venous administration when using the absence of blood aspiration as an indicator of a nonintravascular position of the needle tip (27).
The effect of findings in the previously discussed studies on clinical practice may be that in the preponderance of more recent investigations, an image-guided epidural steroid injection technique will be used. Thus, while there are no analyses findings or other sources of direct evidence to our knowledge, there is a perception based on our clinical practice and the growing number of recent studies that suggests that image-guided epidural steroid injection procedures are increasingly being performed.
The results of the present study should be viewed with a recognition of limitations inherent in the Medicare database and CPT-4 coding system. The Medicare database is composed of aggregated data and has no specific links between the different procedure codes. Some of the codes may overlap, resulting in an inaccurate estimation of procedure volumes. Also, the multilevel codes (with the modifier "51") are included in the total procedure volume; this is mainly a disadvantage in the evaluation of discography, in which three to four disk levels are typically injected. This allows overestimation of the total number of procedures being performed. However, this would most likely have no effect on the data regarding percentage of specialist participation. Nor would it affect reimbursement data, since these numbers represent the total Medicare reimbursement for each code.
However, exclusion of multispecialty groups may influence to a small degree the percentage of provider distribution. The share of spinal injections performed by radiologists, anesthesiologists, surgeons, physiatrists, and other specialists might differ by up to 1% percent from that described. There are also issues regarding coding errors and completeness of the data. Prior to the year 2000, code descriptions were vague, allowing for misinterpretation. Also, the Medicare database represents allowed reimbursement data: If reimbursement for a procedure was denied, it would not appear in the database. During the study period, there was a movement into HMO reimbursement, from 7.4% of benefits in 1993 to 17.1% in 1999. Procedures related to such reimbursement were omitted from the study, but we do not feel that this is likely to affect our results. Although Medicare reimbursement data provide a useful economic perspective, they are not equivalent to the true cost of the procedure. Finally, it should be recognized that Medicare recipients represent an older population; data from this study may not be representative of the volume of spinal injection procedures performed in younger patients.
Future directions for evaluation of volume patterns will become easier because of changes in CPT coding starting in the year 2000. New CPT codes will make it easier to track specific types of epidural and facet joint injections. In addition, there is a new CPT code for vertebroplasty starting in 2001 that will allow investigation of volume patterns for this procedure.
In conclusion, despite an overall increase in percutaneous spinal procedure volume and total reimbursement from 1993 to 1999, nonradiologists are performing most of these procedures. Epidural and facet joint injections are performed almost exclusively by nonradiologists. Radiologists are performing fewer discography procedures than they were previously. With the increasing popularity of intradiskal therapies, such as intradiskal electrothermal modulation, nucleoplasty, and laser discectomy, it is likely that in the years following 1999, anesthesiologists and surgeons will perform more discography procedures and that the trend away from radiologists may continue. Notwithstanding these trends, radiologists have a great deal of experience in placing needles precisely by using image guidance, and therefore, the practice of spinal injection represents a potential area of growth for radiologic practice.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Author contributions: Guarantors of integrity of entire study, J.A.C., W.B.M.; study concepts and design, all authors; literature research, J.A.C., W.B.M.; data acquisition, L.P., J.H.S.; data analysis/interpretation, all authors; manuscript preparation and definition of intellectual content, all authors; manuscript editing, J.A.C., W.B.M., L.P., M.E.S., D.C.L.; manuscript revision/review and final version approval, all authors.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. Fritz, C. Thomas, S. Clasen, C. D. Claussen, J. S. Lewin, and P. L. Pereira Freehand Real-Time MRI-Guided Lumbar Spinal Injection Procedures at 1.5 T: Feasibility, Accuracy, and Safety Am. J. Roentgenol., April 1, 2009; 192(4): W161 - W167. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. K. Freburger, G. M. Holmes, R. P. Agans, A. M. Jackman, J. D. Darter, A. S. Wallace, L. D. Castel, W. D. Kalsbeek, and T. S. Carey The Rising Prevalence of Chronic Low Back Pain Arch Intern Med, February 9, 2009; 169(3): 251 - 258. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. A. Deyo, S. K. Mirza, J. A. Turner, and B. I. Martin Overtreating Chronic Back Pain: Time to Back Off? J Am Board Fam Med, January 1, 2009; 22(1): 62 - 68. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 B. I. Martin, R. A. Deyo, S. K. Mirza, J. A. Turner, B. A. Comstock, W. Hollingworth, and S. D. Sullivan Expenditures and Health Status Among Adults With Back and Neck Problems JAMA, February 13, 2008; 299(6): 656 - 664. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Fritz, T. Niemeyer, S. Clasen, J. Wiskirchen, G. Tepe, B. Kastler, T. Nagele, C. W. Konig, C. D. Claussen, and P. L. Pereira Management of Chronic Low Back Pain: Rationales, Principles, and Targets of Imaging-guided Spinal Injections RadioGraphics, November 1, 2007; 27(6): 1751 - 1771. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 I. A. Young, G. S. Hyman, L. N. Packia-Raj, and A. J. Cole The Use of Lumbar Epidural/Transforaminal Steroids for Managing Spinal Disease J. Am. Acad. Ortho. Surg., April 1, 2007; 15(4): 228 - 238. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. Armon, C. E. Argoff, J. Samuels, and M.-M. Backonja Assessment: Use of epidural steroid injections to treat radicular lumbosacral pain: Report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology Neurology, March 6, 2007; 68(10): 723 - 729. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. Galiano, A. A. Obwegeser, G. Bodner, M. Freund, H. Maurer, F. S. Kamelger, R. Schatzer, and F. Ploner Real-time Sonographic Imaging for Periradicular Injections in the Lumbar Spine: A Sonographic Anatomic Study of a New Technique J. Ultrasound Med., January 1, 2005; 24(1): 33 - 38. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. M. Hooten, M. O. Kinney, and M. A. Huntoon Epidural Abscess and Meningitis After Epidural Corticosteroid Injection Mayo Clin. Proc., May 1, 2004; 79(5): 682 - 686. [Abstract] [PDF]
|
|
|
|
|
|
N. J. Bureau Image-Guided Spine Intervention. Am. J. Roentgenol., August 1, 2003; 181(2): 526 - 526. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| RADIOLOGY | RADIOGRAPHICS | RSNA JOURNALS ONLINE |
