| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
The Practice of Radiology |
1 Department of Diagnostic Imaging, Division of Vascular and Interventional
Radiology, Rhode Island Hospital, 593 Eddy St., Providence, RI 02903.
2 Department of Diagnostic Imaging, Brown Medical School, Providence, RI
02912.
3 Department of Biostatistics, School of Public Health, University of Michigan,
1420 Washington Heights, Ann Arbor, MI 48109-2029.
Received October 17, 2003; accepted after revision March 26, 2004.
Address correspondence to T. P. Murphy
(tmurphy{at}lifespan.org).
OBJECTIVE. Our purpose was to review data from Medicare physician claims to ascertain differences in annual volumes of renal artery interventions versus surgery and differences in utilization of renal artery interventions among geographic regions according to physician provider type between 1996 and 2000.
MATERIALS AND METHODS. We analyzed claims submitted to Medicare in 1996, 1998, and 2000 and extracted claims for renal artery angioplasty, stent placement, or bypass surgery. Analyses were performed for percutaneous renal artery interventions categorized by Centers for Medicare & Medicaid Services (CMS) geographic region and physician provider type.
RESULTS. Between 1996 and 2000, the total volume of renal revascularization (surgical and percutaneous) increased 62%, from 13,380 to 21,660 procedures. The annual volume of renal artery surgery decreased 45% in 2000, compared with the volume in 1996. Annual volumes of renal artery angioplasty and stent placement increased 2.4-fold in 2000 compared with those in 1996. Most growth in percutaneous renal artery interventions is attributed to added provision by cardiologists, who increased their annual volume 3.9-fold. More than a threefold difference in rates of use of renal artery interventional procedures across CMS regions was found. In the Southeast region, the volume of renal artery interventions by cardiologists increased more than 15-fold.
CONCLUSION. Among Medicare beneficiaries, the volume of percutaneous renal artery interventions is increasing rapidly, whereas the volume of renal artery surgery is declining. Most growth in percutaneous renal artery revascularization is attributed to increased performance by cardiologists; explosive growth in annual procedure volume by cardiologists occurred in some regions. Marked disparity in use among CMS regions was found.
Most patients with hypertension have essential hypertension, with only 5–10% having secondary hypertension as a result of causes such as renal artery stenosis, pheochromocytoma, or aldosteronoma. Renal artery stenosis is common, seen in 6–18% of those patients undergoing coronary arteriography [1–3] and in 16–40% of those patients undergoing aortography for aneurysms or peripheral vascular disease [4–10]. Typical indications for revascularization of stenotic renal arteries include hypertension and chronic renal insufficiency [11, 12]. However, renal artery interventions are controversial. Three randomized clinical trials of renal artery angioplasty compared with medical therapy showed no convincing benefit of angioplasty with regard to blood pressure control [13–15]. The purpose of this investigation was to review use by Medicare beneficiaries of renal artery revascularization between 1996 and 2000 to determine the change in the annual volume of renal artery interventions and renal artery surgery, to compare rates of use by geographic region, to compare volume by physician specialty type performing renal artery interventions and the change in volume by specialty mix over time, and to assess the change in volume by physician specialty in geographic regions over time.
Materials and Methods
We performed an analysis of data from the Centers for Medicare & Medicaid Services (CMS) for 3 years (1996, 1998, and 2000), covering a span of 5 years. The files used were the provider or Part B 5% files. These files contain data from a random 5% sample of all claims submitted by physicians for reimbursement. The data were obtained by the Lewin Group (Falls Church, VA) on IBM 3480 tape cartridges and processed on an IBM 9672-R24 mainframe computer (International Business Machines) using Statistical Analysis Software (SAS Institute). Each file contains information for roughly 1.9 million beneficiaries and 30 million claims. The 5% files were used because they were likely to be statistically valid, were considerably less expensive than 100% files, and are commonly used in this type of analysis. We cross-checked our results against the Part B procedure summary file for Current Procedural Terminology (CPT) [16] codes of interest to ascertain the validity of extrapolating results from the 5% file for 1998. This audit showed good consistency, with the extrapolated total for code 37205 within 2% of the actual total from the 100% file for that year and extrapolated total for 35471 within 7% of the 100% file total. To estimate totals for the entire CMS population, we multiplied the results from 5% files by 20. Institutional review board approval was obtained and informed consent was waived.
The 5% data set was examined for claims that included CPT codes indicating potential renal intervention procedures (35471—transluminal balloon angioplasty, percutaneous; renal or visceral artery; and 37205—transcatheter placement of an intravascular stent[s]; [non-coronary vessel], percutaneous; [initial vessel]) and surgical CPT codes (35341—thromboendarterectomy, mesenteric, celiac, or renal; 35536—bypass, splenorenal; 35560—bypass, aortorenal; 35631—bypass, aortoceliac, aortomesenteric, aortorenal; 35636—bypass, splenorenal). CPT codes 37205 and 35741 are not specific to anatomic site; therefore, analysis of other codes that appeared on the same claim as codes 35471 and 37205 was performed to ascertain the artery in which the intervention was performed (Tables 1, 2, 3). This analysis included consideration of other procedure codes and principal diagnosis codes that appeared on the claims. International Classification of Disease, 9th edition (ICD-9) codes [17] were selected as possibly indicating a renal artery procedure by searching a commercially available code database (PointChart, Boston Medical Systems) using the following terms: hypertension, renal failure, and heart failure (Tables 1, 2, 3). The details of the rules used to distinguish claims that represented renal artery interventional procedures from those that represented interventions in other vascular territories are presented as part of the Results section (Tables 1, 2, 3). Surgical CPT codes are relatively site-specific and were taken to indicate renal artery interventions when they were submitted. Data were tabulated using the patient encounter as the primary data element. If more than one CPT code of interest (such as a balloon angioplasty code and a stent code or a bilateral procedure) appeared on a claim for a patient for the same day of service, this information was counted as one patient encounter.
|
|
|
To validate the assumptions made regarding which claims represented renal artery interventions, we analyzed a sample of 52 bills from the year 2000 from five sites throughout the United States, including California, New York, New Hampshire, Virginia, and Rhode Island. These sites were asked to submit CPT and ICD-9 codes used by their practices on claims known to be from renal artery angioplasty or stent procedures and for services known not to represent renal artery angioplasty or stent placement, without any patient identifying information. For this sample of 52 bills, 26 renal artery procedures and 26 "not" renal artery stent procedures (including renal arteriography, iliac artery stenting, venous stenting, and aortic endograft placement) were submitted.
Results obtained from the 5% files were presented as estimated Medicare annual volumes by multiplying by 20. The frequency of service per 100,000 beneficiaries was compared in each of 10 CMS geographic regions, and these rates were compared over time for each CMS region (and for all Medicare beneficiaries). The extrapolated volumes in each year were divided according to physician specialty type to evaluate provider specialty trends over time, and provider specialty trends over time within geographic regions were also evaluated. All volume data reported are the extrapolated total use estimates for the entire Medicare population, in which the numbers from 5% samples were multiplied by 20. For all reporting and comparisons, descriptive statistics and tabulations were used. No statistical testing was performed; thus, no statistical significance such as p value was reported because the primary data of renal intervention encounters were not directly observed but rather were extracted and estimated by applying an algorithm to arrive at a subset of what we thought best fit such encounters.
Results
Validation of Rules Used to Identify Percutaneous Renal Artery Interventions
For our sample of 52 (26 renal and 26 nonrenal) known cases used to
evaluate the accuracy of our rules for analyzing claims, 32 included CPT code
37205 (stent first order vessel) but not code 35471 (renal or visceral
angioplasty), six that included CPT code 35471 and not 37205, and seven that
contained both. Because of the relative lack of anatomic site specificity of
37205 compared with 35471 (which comprises only renal and visceral
angioplasty, the latter being a relatively small number), claims with code
37205 (the code used for peripheral artery stent placement) but not 35471 were
considered to be the most challenging to ascertain anatomic site. For all the
sample claims, our rules were 80% sensitive and 96% specific for identifying
percutaneous renal artery revascularization procedures in our sample
population. For the purpose of examining variations across geographic regions
and physician provider types, it was desirable to have optimum specificity so
that only renal artery interventions were included for the analysis. However,
our cumulative results of annual procedure volumes are likely to be
underestimated because of the reduced sensitivity of our algorithm.
Nevertheless, the analysis of trends over time should be valid because no
systematic change in coding in Medicare claims submission occurred over the
5-year period.
Annual Renal Revascularization Volume and Rate Trends
Results extrapolated to the Medicare population as a whole indicate that
approximately 7,660 claims for renal artery interventional procedures (balloon
angioplasty or stent or both) were submitted in 1996, compared with 11,400 in
1998 and 18,520 in 2000 (Table
4). In the 5-year interval, the annual volume of Medicare
beneficiaries who had claims submitted for renal artery interventional
procedures increased 2.4-fold. In that time, the volume of claims for renal
artery angioplasty alone (CPT code 35471) without stents grew only 16%,
whereas the number of claims for renal artery stent placement without
angioplasty charges grew 3.7-fold, and the annual volume of claims including
charges for both renal artery angioplasty and renal artery stent placement
grew 2.2-fold. Use rates per 100,000 beneficiaries increased from 25 to 61,
also a 2.4-fold increase.
|
Contrary to the rise in volume of claims for percutaneous revascularization, annual claims for surgical revascularization procedures decreased substantially in 2000 compared with claims submitted in 1996. Although 5,720 such services were charged in 1996, the number decreased to 3,740 in 1998 and to 3,140 in 2000, a decrease in annual volume of 45% over the 5-year period. Although among Medicare beneficiaries, the annual increase in percutaneous revascularization of roughly 10,680 procedures from 1996 to 2000 cannot be attributed to a shift from surgical to percutaneous revascularization alone, the relative proportion of surgical revascularization in the estimated total renal artery revascularization procedures showed significant decline over time: 43% (5,720/13,380) in 1996, 25% (3,740/15,140) in 1998, and 14% (3,140/21,660) in 2000. The total volume of renal revascularization procedures (surgical and percutaneous) increased 62% from 13,380 to 21,660 in the same 5-year period.
Geographic Variation in Renal Artery Angioplasty and Stent Placement
For all years examined, a broad range in use rates of renal artery
interventional procedures by CMS region
(Table 5) was seen. For 2000,
61 claims were submitted per 100,000 CMS beneficiaries. Use rates ranged from
26 renal artery interventions per 100,000 beneficiaries in the Northeast
(Connecticut, Maine, Massachusetts, New Hampshire, New York, Rhode Island, and
Vermont) to 87 per 100,000 in the South (Florida and Georgia), a more than
three-fold difference (Table
5). These data are from Medicare beneficiaries, who are mostly
older than 65 years, so differences in ages of the populations across
geographic regions should be small and should not explain the discrepancies.
The greatest absolute increase in the rate of use of renal artery
interventions between 1996 and 2000 occurred in the Southeast (Alabama,
Kentucky, Mississippi, North Carolina, South Carolina, and Tennessee), with an
increase from 22 per 100,000 beneficiaries to 86 per 100,000 beneficiaries.
Examining relative increases in use rates over time by region, we found that
the increase in annual volume over this time period ranged from 1.4 in the
Northeast region to 4.8 in the Keystone region. Six regions (Keystone,
Southeast, South, Great Lakes, Pacific, and Southwest) increased use rates per
beneficiary by more than twofold between those years.
|
Provision of Renal Artery Intervention Services by Provider Specialty Type
In each of the years, we could identify provider types as cardiology,
radiology, or surgery for between 95% and 96% of the total renal percutaneous
revascularization claims (Table
6). Significant trends were observed with regard to physician
provider type. In 1996, "radiologists" ("radiology"
and "interventional radiology") performed most renal artery
angioplasties and stent placements that were identifiable to a provider type
for Medicare beneficiaries, totaling 4,700 services, compared with 2,380 for
"cardiologists" ("cardiology" and "internal
medicine") (Table 6). By
2000, cardiologists submitted charges for 9,220 patient encounters of renal
artery angioplasty or stent placement, an increase in annual volume of
3.9-fold from 1996. By contrast, radiologists increased their performance of
this service to 7,660 in 2000, an increase of 1.6-fold in annual volume
compared with the volume in 1996. Comparing the same two time periods, we
found that surgeons increased their volume of percutaneous renal artery
revascularization claims 1.5-fold but accounted for only 4% of the total in
2000, or 760 claims.
|
Provider Specialty-Type Trends Within Geographic Regions
Six of 10 regions increased use rates of renal artery interventions by more
than twofold between 1996 and 2000. In these six regions, the number of claims
for renal artery interventions submitted by cardiologists increased from 1,920
to 7,560. In some regions, explosive growth in annual procedure volumes by
cardiologists occurred (Table
7). For example, the annual volume of renal artery interventions
by cardiologists increased more than 15-fold in the Southeast region, almost
eightfold in the Keystone region, more than sixfold in the Pacific region, and
more than fivefold in the South region
(Table 7).
|
Discussion
Physician practice patterns are known to vary widely for a number of diseases and procedures. For example, rates of use of coronary artery bypass surgery in Medicare beneficiaries ranged from seven per 10,000 to 23 per 10,000 in 13 geographic regions in 1981 [18]. In this same study, rates of carotid endarterectomy ranged from six per 10,000 to 23 per 10,000, and the use of upper endoscopy ranged from 94 per 10,000 to 153 per 10,000 [18]. Of 123 procedures examined, 67 had at least a threefold difference in use among geographic regions [18]. Regional differences have also been found in the use of beneficial and detrimental cardiac medications and in the use of coronary angiography, angioplasty, and coronary artery bypass surgery [19, 20]. Variability in health care delivery for a particular disease has been attributed to a lack of medical consensus regarding therapy [19], but even when practice guidelines are published, variability is still common [21–24].
The annual rate of renal artery interventional procedures billed to Medicare increased 2.4-fold in 2000 compared with 1996, and the annual volume of renal artery bypass surgery decreased by 45% during the same time period. We found that delivery of care for beneficiaries with renal artery stenosis is highly dependent on geographic region and on the specialty of the provider.
One limitation of our analysis is that for nonsurgical revascularization
claims, we had to apply an algorithm to arrive at claims that were specific to
renal revascularization. When validated against known cases, this algorithm
was shown to have 80% sensitivity and 96% specificity. The relatively low
sensitivity means that our annual estimates for the volume of nonsurgical
renal revascularization procedures may have been underestimated. The
substantial proportion of migration of Medicare beneficiaries to managed care
between 1996 and 2000 (
7%, Haught R, Lewin Group, personal communication)
also results in for further underestimation of the volume for 2000 because of
the decreased number of beneficiaries included in our 5% sample. Even if the
bias may be of concern, it should not change our conclusion because the bias
would be in the direction of underestimation in the total volume change in
renal artery interventional procedures, whereas our results support a
substantial increase in volume. Practice pattern variation is of interest to
health care policy makers because it points to either lack of consensus
regarding disease management, lack of availability of some services in some
regions or at nonteaching hospitals, or potential overuse or abuse of some
services. For some procedures that have been identified as subject to
geographic variation, such as carotid endarterectomy, coronary angiography,
and upper endoscopy, the variability has been attributed to unjustified
overutilization [25]. However,
generalizations about the cause for variations in physician practices usually
cannot be made from these data alone
[26]. According to Chassin
[26], "We do not know
whether physicians in high-use areas performed too many procedures, whether
physicians in low-use areas performed too few, or whether neither or both of
these explanations are accurate." Radiologists increased their annual
volume of renal artery interventional services by 1.6-fold between 1996 and
2000, whereas cardiologists increased theirs by 3.9-fold. In five of 10 CMS
regions, annual procedure volume by internists and cardiologists grew by at
least a factor of 5 and in one, the Southeast, by a factor of 15. The growth
in renal artery interventional procedures by cardiologists is not surprising
given their recent interest in performing procedures in vascular beds other
than the heart. Renal artery stenosis has received considerable interest in
the cardiology literature since the mid-1990s, when cardiologists began the
practice of "drive-by" renal angiography during coronary
angiography [27] and found
that renal artery stenosis is present in 6–18% of those who undergo
coronary angiography
[1–3].
Some Medicare carriers developed policies denying reimbursement for renal
artery procedures performed on the same day of service as coronary angiography
if accepted indications were not present.
There is true clinical equipoise regarding decisions for treating those with renal artery stenosis and hypertension; many practitioners treat such patients medically, whereas in other centers, renal artery angioplasty is common. Detecting a difference between medical therapy and renal artery angioplasty for high blood pressure in randomized trials has been elusive [13–15]. Nevertheless, percutaneous renal artery revascularization is experiencing considerable growth. It has been stated that practice guidelines are the best hope for improving the quality of care [28, 29]. Practice guidelines require evidence based on scientific clinical trials or, in their absence, consensus opinions. No consensus exists on how to best treat patients with hypertension and renal artery stenosis, and a large, randomized clinical trial is needed to provide information from which practice guidelines could be eventually developed.
Acknowledgments
We thank Randy Haught, senior scientist at the Lewin Group, for his invaluable contributions to this work and Elizabeth Hann for assistance with manuscript preparation.
References
CiteULike 
Complore 
Connotea 
Del.icio.us 
Digg 
Reddit 
Technorati What's this?
This article has been cited by other articles:
|
|
 |
|
  J. A. Bittl Damage Control for Renal Artery Stenting Circulation, May 27, 2008; 117(21): 2724 - 2726. [Full Text] [PDF]
|
|
|
|
 |
|
 G. J Dubel and T. P Murphy The role of percutaneous revascularization for renal artery stenosis Vascular Medicine, May 1, 2008; 13(2): 141 - 156. [Abstract] [PDF]
|
|
|
|
 |
|
 S. C. Textor Atherosclerotic Renal Artery Stenosis: Overtreated but Underrated? J. Am. Soc. Nephrol., April 1, 2008; 19(4): 656 - 659. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. W Dear, P. L Padfield, and D. J Webb New guidelines for drive-by renal arteriography may lead to an unjustifiable increase in percutaneous intervention Heart, December 1, 2007; 93(12): 1528 - 1532. [Full Text] [PDF]
|
|
|
|
 |
|
 C. M. Johnson and K. J. Hodgson The Role of Embolic Protection Devices in Renal Artery Stenting Perspectives in Vascular Surgery and Endovascular Therapy, September 1, 2007; 19(3): 266 - 271. [Abstract] [PDF]
|
|
|
|
|
|
C. J. Cooper and T. P. Murphy The Case for Renal Artery Stenting for Treatment of Renal Artery Stenosis Circulation, January 16, 2007; 115(2): 263 - 270. [Full Text] [PDF]
|
|
|
|
 |
|
 E. Balk, G. Raman, M. Chung, S. Ip, A. Tatsioni, A. Alonso, P. Chew, S. J. Gilbert, and J. Lau Effectiveness of Management Strategies for Renal Artery Stenosis: A Systematic Review Ann Intern Med, December 19, 2006; 145(12): 901 - 912. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. J. White Catheter-Based Therapy for Atherosclerotic Renal Artery Stenosis Circulation, March 21, 2006; 113(11): 1464 - 1473. [Full Text] [PDF]
|
|
|
|
|
|
V. D. Garovic and S. C. Textor Renovascular Hypertension and Ischemic Nephropathy Circulation, August 30, 2005; 112(9): 1362 - 1374. [Full Text] [PDF]
|
|
|
|
|
|
J. A. Silva, A. W. Chan, C. J. White, T. J. Collins, J. S. Jenkins, J. P. Reilly, and S. R. Ramee Elevated Brain Natriuretic Peptide Predicts Blood Pressure Response After Stent Revascularization in Patients With Renal Artery Stenosis Circulation, January 25, 2005; 111(3): 328 - 333. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
