I

Itsrig

DU

MRP

8

Trends in the Utilization of CTAngiography and MR Angiography ofthe Head and Neck in the Medicare

PopulationDavid P. Friedman, MD, David C. Levin, MD, Vijay M. Rao, MD

Purpose: The aim of this study was to analyze trends in the utilization of CT angiography (CTA) and MRangiography (MRA) of the head and neck in the Medicare population over a 6-year interval.

Methods: Nationwide Medicare Part B fee-for-service databases were reviewed. Current Procedural Termi-nology® codes for CTA and MRA of the head and neck were selected. MRA codes included studies withoutcontrast, with contrast, and without and with contrast. Yearly and aggregate procedure volumes were comparedfor each Current Procedural Terminology code and modality. Data were also analyzed regarding contrastutilization and cost.

Results: From 2002 to 2007, the volume of head CTA increased by 827%, and the overall volume of headMRA increased by 39%. The year-to-year percentage increase in overall volume of head MRA declinedthroughout the study period; almost all of the increase in the overall volume of head MRA occurred from 2002to 2005. The volume of neck CTA increased by 1,074%, and the overall volume of neck MRA increased by31%. An 18% decrease in the volume of neck MRA without contrast was offset by a 104% increase in thevolume of neck MRA using contrast. The year-to-year percentage increase in the overall volume of neck MRAdeclined from 2002 to 2005; there was a decrease in volume of 3% from 2005 to 2007. From 2002 to 2007,when considering all study types, procedure volume increased by 71%; aggregate allowable charges increased by$181 million. Examinations using contrast increased by 235%. In 2002, 23% of examinations used contrast; in2007, 46% of examinations used contrast.

Conclusions: The rate of growth for head and neck CTA was dramatically higher than for MRA. Neck MRAusing contrast also showed substantial growth. The Medicare population is now receiving more contrastmaterial and radiation to noninvasively assess the arterial vasculature of the head and neck.

Key Words: Utilization, CT angiography, MR angiography

J Am Coll Radiol 2010;7:854-858. Copyright © 2010 American College of Radiology

pcatsmid

Ct6t

NTRODUCTION

n the past 2 decades, there has been a steady increase inhe utilization of high-cost, advanced imaging studiesuch as CT and MRI in the United States [1-4]. Theeasons for increased utilization are multifactorial andnclude technologic advances, changes in best practiceuidelines, self-referral, the aging population, patient ex-

epartment of Radiology, Jefferson Medical College and Thomas Jeffersonniversity Hospital, Philadelphia, Pennsylvania.

Corresponding author and reprints: David P. Friedman, MD, Jeffersonedical College and Thomas Jefferson University Hospital, Department of

adiology, 132 South 10th Street, Suite 1068, Main Building, Philadelphia

tA 19107; e-mail: david.friedman@jefferson.edu.

54

ectations, and medicolegal concerns. The escalatingost of diagnostic imaging has made national headlines,s has concern related to increasing radiation exposure tohe general population of the United States [5-7]. Expo-ure of the population to ionizing medical radiation grewore than 7-fold between the early 1980s and 2006, an

ncrease attributable largely to the dramatic growth in theiagnostic capabilities and use of CT [8-10].In this study, we analyzed trends in the utilization of

T angiography (CTA) and MR angiography (MRA) ofhe head and neck in the Medicare population over a-year interval. To provide further perspective regardinghese trends, procedure volumes for duplex ultrasound of

he carotid arteries, as well as conventional craniocervical

© 2010 American College of Radiology0091-2182/10/$36.00 ● DOI 10.1016/j.jacr.2010.05.007

craviereerc

M

Uttpctb

csanc7anwdg

ts7(gc

R

IM3civr

irDotHaph2mT

ipicidwn7aabotd

8,2

Friedman et al/Utilization of CTA and MRA 855

arotid and vertebral angiography, were also analyzed. Byestricting the study to a particular patient populationnd a small number of procedures evaluating the sameascular anatomy, we hoped to gain a better understand-ng of trends in the utilization of these procedures. Forxample, does the growth in volume of one procedureesult in a corresponding decline in another procedurevaluating the same anatomy? Is the population of inter-st receiving more or less contrast material? More or lessadiation? What impact do these factors have on overallosts?

ETHODS

sing the Physician Supplier Procedure Summary Mas-er Files, nationwide Medicare Part B fee-for-service da-abases from 2002 to 2007 were reviewed. All studieserformed by radiologists, nonradiologists, multispe-ialty groups, and independent diagnostic testing facili-ies on fixed, fee-for-service Medicare beneficiaries inoth inpatient and outpatient settings were included.The Current Procedural Terminology® (CPT®)

odes for CTA and MRA of the head and neck wereelected; a total of 8 codes were evaluated. The CTngiographic codes included head CTA (70496) andeck CTA (70498). The MR angiographic codes in-luded studies without contrast (head, 70544; neck,0547), studies with contrast (head, 70545; neck, 70548),nd studies without and with contrast (head, 70546;eck, 70549). Yearly and aggregate procedure volumesere compared for each CPT code, as well as each mo-ality (CTA vs MRA). The data were also analyzed re-arding contrast utilization and cost.

Aggregate procedure volumes for duplex ultrasound ofhe carotid arteries (bilateral, 93880; unilateral or limitedtudy, 93882), cranial carotid angiography (unilateral,5665; bilateral, 75671), cervical carotid angiographyunilateral, 75676; bilateral, 75680), and vertebral an-iography (cervical or intracranial, 75685) were also

Table 1. Yearly procedure volumes of head CTA an

Study Type 2002 2003 2Head CTA 8,987 16,037 2Head MRA without

contrast255,643 290,443 32

Head MRA withcontrast

6,653 6,666

Head MRA without andwith contrast

10,091 11,693 1

All head MRA 272,387 308,802 34

ompared (2002 vs 2007). u

ESULTS

n 2002, there were 34,997,612 fixed, fee-for-serviceedicare beneficiaries; this number increased to

5,434,519 in 2007 (�1%). Because of the minimalhange in the number of fee-for-service beneficiaries dur-ng the study period, percentage changes in procedureolumes almost exactly parallel percentage changes in theate of procedures per 100,000 beneficiaries.

Comparing 2002 with 2007, the volume of head CTAncreased steadily from 8,987 to 83,297 (�827%); theate per 100,000 beneficiaries increased from 26 to 235.uring this time, the overall volume of head MRA (with-

ut contrast, with contrast, and without and with con-rast) increased from 272,387 to 377,820 (�39%).owever, the year-to-year percentage increase in the over-

ll volume of head MRA declined throughout the studyeriod, and almost all of the increase in the overall volume ofead MRA occurred from 2002 to 2005. From 2002 to007, the overall volume of all head examinations (bothodalities) increased from 281,374 to 461,117 (�64%).hese data are summarized in Table 1.Comparing 2002 with 2007, the volume of neck CTA

ncreased from 9,796 to 115,021 (�1,074%); the rateer 100,000 beneficiaries increased from 28 to 325. Dur-ng this time, the overall volume of neck MRA (withoutontrast, with contrast, and without and with contrast)ncreased from 192,653 to 253,170 (�31%). An 18%ecrease in the volume of neck MRA without contrastas more than offset by a 104% increase in volume ofeck MRA using contrast (CPT codes 70548 and0549). The year-to-year percentage change in the over-ll volume of neck MRA declined from 2002 to 2005,nd the overall volume of neck MRA actually decreasedy 3% from 2005 to 2007. From 2002 to 2007, theverall volume of all neck examinations (both modali-ies) increased from 202,449 to 368,191 (�82%). Theseata are summarized in Table 2.Comparing 2002 with 2007, the volume of duplex

MRA from 2002 to 2007

Year %Change

2002-20074 2005 2006 200780 47,757 66,717 83,297 827%77 350,328 355,355 355,406 39%

68 6,489 6,766 6,200 �7%

74 15,842 16,457 16,214 61%

19 372,659 378,578 377,820 39%

d

009,87,2

6,5

4,3

ltrasound of the carotid arteries (two CPT codes) in-

cig5r(r(g2

t(pC

m

Ce(

cI4m

cna$aiwc(

D

Daaosnh2CHpouugpwjottop

24

856 Journal of the American College of Radiology/Vol. 7 No. 11 November 2010

reased from 2,533,820 to 3,038,905 (�20%). Compar-ng 2002 with 2007, the volume of cranial carotid an-iography (two CPT codes) decreased from 84,060 to3,747 (�36%); the volume of cervical carotid angiog-aphy (two CPT codes) decreased from 92,052 to 58,389�37%); the volume of craniocervical vertebral angiog-aphy (one CPT code) decreased from 58,048 to 46,870�19%). When considering all conventional angio-raphic procedures, the overall volume decreased from34,160 to 159,006 (�32%).When considering all CT angiographic procedures,

he overall volume increased from 18,783 to 198,318�956%). The number of conventional angiographicrocedures decreased by 75,154, while the number ofT angiographic procedures increased by 179,535.Aggregate procedure volumes (2002 vs 2007) are sum-arized in Table 3.Comparing 2002 with 2007, when considering both

TA and MRA, and all study types, the number ofxaminations increased from 483,823 to 829,308�71%). However, the number of examinations using

Table 2. Yearly procedure volumes of neck CTA an

Study Type 2002 2003Neck CTA 9,796 18,882Neck MRA without contrast 114,566 112,107Neck MRA with contrast 26,374 37,956Neck MRA without and

with contrast51,713 71,657

All neck MRA 192,653 221,720

Table 3. Aggregate procedure volumes, 2002 vs2007

Study Type

Year %Change

2002-20072002 2007Head CTA 8,987 83,297 827%All head MRA 272,387 377,820 39%All head

examinations281,374 461,117 64%

Neck CTA 9,796 115,021 1,074%All neck MRA 192,653 253,170 31%All neck

examinations202,449 368,191 82%

All CTA 18,783 198,318 956%Duplex

ultrasound2,533,820 3,038,905 20%

Catheter 234,160 159,006 �32%

cangiography

ontrast increased from 113,614 to 380,410 (�235%).n 2002, 23% of examinations used contrast; in 2007,6% of examinations used contrast. These data are sum-arized in Table 4.Comparing 2002 with 2007, Medicare allowable

harges (including technical and professional compo-ents, in 2009 dollars; 80% of the total paid by Medicarend 20% paid by patients or coinsurance) increased from251 million to $432 million (�72%). However, allow-ble charges for studies without contrast (2 CPT codes)ncreased from $178 million to $212 million (�19%),hile allowable charges for studies using contrast (6 CPT

odes) increased from $72 million to $219 million�203%). These data are summarized in Table 5.

ISCUSSION

uring the study interval, the rate of growth of headnd neck CTA was dramatically higher than for headnd neck MRA, and there was minimal change in theverall MRA volume from 2005 to 2007. These datauggest that head and neck CTA is replacing head andeck MRA for many patients. However, the volume ofead and neck MRA increased by �10% per year from002 to 2005, and the rate of growth of head and neckTA increased by �20% per year from 2005 to 2007.ence, although one modality (CTA) seems to be

artly replacing another (MRA), growth in the volumef neurovascular imaging in the Medicare populationsing costly, advanced imaging modalities continuednabated. Comparing 2002 with 2007, there was 71%rowth in the volume of all CT and MR angiographicrocedures, and a commensurate 72% increase in cost,hich amounted to a cost increase of $181 million in

ust 6 years for interrogating only neurovascular anat-my with these modalities in the Medicare popula-ion. Of note, these figures do not include the addi-ional cost associated with the 20% increase in volumef duplex ultrasound of the carotid arteries, nor theotential cost savings associated with the decrease in

MRA from 2002 to 2007

Year %Change

2002-2007004 2005 2006 20077,482 63,030 93,230 115,021 1074%8,968 101,678 91,718 93,492 �18%7,996 51,632 55,869 50,780 93%2,977 108,904 118,199 108,898 111%

9,941 262,214 265,786 253,170 31%

d

23

1049

onventional angiographic procedures.

ngtvtwrtaOdsmtptaCtstn

tehwatEsntt

gogtpttc�(clttbnailrommrMydn

tupcpf

3,6

Friedman et al/Utilization of CTA and MRA 857

Technical advancements in multidetector CT scan-ers have undoubtedly contributed to the dramaticrowth in CTA of the head and neck and the decline inhe volume of conventional craniocervical carotid andertebral angiography; there were 75,154 fewer conven-ional angiographic studies performed in 2007 comparedith 2002. Indeed, for many indications, clinicians and

adiologists now regard CTA as a surrogate for conven-ional angiography; from the perspective of patient safetynd comfort, this represents a tremendous improvement.f note, our database does not enable tracking of proce-

ures for individual beneficiaries; hence, our study as-essed procedure volumes, not patients. With this inind, even if it is assumed that the 75,154 fewer conven-

ional angiographic procedures performed were all re-laced by CTA (and if this number is subtracted from theotal number of CT angiographic studies), there was stillmarked increase of 104,381 (�556%) in the volume ofTA of the head and neck from 2002 to 2007. Clearly,

he majority of the growth in CTA is unrelated to itsubstitution for conventional angiography; anecdotally,his corroborates the experience in our own academiceuroradiology practice.Comparing 2002 with 2007, regarding neck MRA,

here was a strong trend toward the use of contrast-nhanced studies; moreover, these studies typically use aigh volume of gadolinium-based contrast agents. Alongith the dramatic growth of head and neck CTA, this

ccounts for the increase of 235% in studies using con-rast (and the commensurate increase of 203% in cost).ven if the 75,154 fewer conventional angiographic

tudies performed in 2007 are deducted from the totalumber of CT and MR angiographic studies using con-rast, there was still an increase in 169%. Of note, con-rast loads for neurovascular CTA and conventional an-

Table 4. Utilization of contrast, 2002 vs 2007

Study Type 2All CTA/MRA 48Without contrast only 37With, and without and with contrast 11

Table 5. Medicare allowable charges,� 2002 vs 200

Study TypeAll CTA/MRAWithout contrast only (2 codes)With, and without and with contrast (6 codes)

�Includes technical and professional components, in 2009 dollars.

iography are generally similar. In 2002, approximatelyne-quarter of all neurovascular CT and MR angio-raphic studies used contrast; in 2007, this had increasedo approximately one-half. Clearly, the elderly Medicareopulation is now receiving markedly more contrast ma-erial to noninvasively assess the arterial vasculature ofhe head and neck. Of note, age is a key predictor ofhronic kidney disease, and 11% of individuals aged65 years without hypertension or diabetes have stage 3

glomerular filtration rate, 30-59 mL/min), or worse,hronic kidney disease [11]. Although serum creatinineevels are now routinely checked in elderly patients beforehe administration of contrast (representing an addi-ional, indirect cost per imaging study), this contrasturden (and potential for nephrotoxicity or theoreticallyephrogenic systemic fibrosis) cannot be viewed as desir-ble. Similarly, the growth of CTA is accompanied by anncrease in ionizing radiation; although this is somewhatess of a concern in the elderly population, this increase inadiation also cannot be viewed as desirable. Althoughur results cannot necessarily be generalized to the re-ainder of the population, the typical indications forany of these studies (eg, evaluate for intracranial aneu-

ysm, carotid stenosis) are applicable to patients belowedicare age. As such, it is reasonable to assume that

ounger patients are also increasingly receiving more ra-iation to assess the arterial vasculature of the head andeck.Screening of the cervical carotid arteries for stenosis is

ypically performed using duplex ultrasound, and the vol-me of these studies increased by 20% during the studyeriod. An increase in these screening studies undoubtedlyontributed to the growth of CTA and MRA, as someatients will require corroboration of ultrasound findings orurther anatomic evaluation. Technological advances in the

Year % Change2002 to 20072 2007

23 829,308 71%09 448,898 21%14 380,410 235%

Year %Change

2002-20072002 2007251,420,187 $431,989,057 72%178,893,562 $212,515,285 19%$72,526,625 $219,473,772 203%

003,80,2

7

$$

tucbaiaansca

qcitmdtfwfipadHeddrtmtab

R

1

1

1

1

1

1

1

858 Journal of the American College of Radiology/Vol. 7 No. 11 November 2010

reatment of neurovascular diseases may also have contrib-ted to the growth of CTA and MRA; for example, a Medi-are patient not regarded as a surgical candidate might nowe amenable to therapy with a newer, minimally invasivengiographic procedure (eg, aneurysm coiling), necessitat-ng more noninvasive imaging before treatment. CTA islso being increasingly utilized in the evaluation of hyper-cute infarction. Regardless of the causes of the growth ofoninvasive procedures, given the stability in size of thetudy population, it is clear that a similar number of Medi-are patients is now being imaged more frequently, with anttendant increase in cost.

These data do not provide answers to many importantuestions. For example, we cannot determine if the in-rease in advanced neurovascular imaging has resulted inmprovement in the outcomes of individual patients orhe health of the Medicare population as a whole. Howany fewer patients avoided complications related to the

ecrease in conventional angiography? How many addi-ional patients had complications related to the morerequent administration of contrast material? How oftenere clinically important (or unimportant) incidentalndings detected that required further evaluation? Com-ared with conventional angiography, CTA and MRAre indeed “noninvasive”; does this lead to more redun-ant or more frequent follow-up studies than necessary?ow often are CTA and MRA used for screening of the

xtracranial carotid and vertebral arteries, rather thanuplex ultrasound? What are the clinical implications ofetecting certain neurovascular lesions (eg, small aneu-ysms) in the elderly population? The answers to many ofhese questions fall into the realm of evidence-basededicine, and they are extremely important [12-14]. In

he current economic and health care climate, insurersnd politicians will certainly want to know what gains areeing achieved for any increase in cost [15,16].

EFERENCES

1. Medicare Payment Advisory Commission. Report to the Congress: Medi-care payment policy. Available at: http://www.medpac.gov/documents/

Mar07_EntireReport.pdf. Accessed August 12, 2010.

2. Maitino AJ, Levin DC, Parker L, Rao VM, Sunshine JH. Nationwidetrends in rates of utilization of noninvasive diagnostic imaging among theMedicare population between 1993 and 1999. Radiology 2003;227:113-7.

3. Bhargavan M, Sunshine JH. Utilization of radiology services in theUnited States: levels and trends in modalities, regions, and populations.Radiology 2005;234:824-32.

4. Pennsylvania Health Care Cost Containment Council. The growth indiagnostic imaging utilization. Harrisburg: Pennsylvania Health CareCost Containment Council; 2004.

5. Landro L. Radiation risks prompt push to curb CT scans. The Wall StreetJournal. March 2, 2010. Available at: http://online.wsj.com/article/NA_WSJ_PUB:SB10001424052748704299804575095502744095926.html.Accessed August 12, 2010.

6. Berenson A. Study finds radiation risk for patients. The New York Times.August 26, 2009. Available at: http://www.nytimes.com/2009/08/27/health/research/27scan.html. Accessed August 12, 2010.

7. Tanner L. Experts say even Obama getting too many med tests. Availableat: http://abcnews.go.com/Health/wireStory?id�10081845. AccessedAugust 12, 2010.

8. Brenner DJ, Hall EJ. Computed tomography—an increasing source ofradiation exposure. N Engl J Med 2007;357:2277-84.

9. Frush DP, Applegate K. Computed tomography and radiation: under-standing the issues. J Am Coll Radiol 2004;1:113-9.

0. National Council on Radiation Protection and Measurements. Ionizingradiation exposure of the population of the United States (report 160).Bethesda, Md: National Council on Radiation Protection and Measure-ments; 2009.

1. Coresh J, Astor BC, Greene T, Eknoyan G, Levey AS. Prevalence ofchronic kidney disease and decreased kidney function in the adult USpopulation: third national health and nutrition examination survey. Am JKidney Dis 2003;41:1-12.

2. Gazelle GS, McMahon PM, Siebert U, Beinfeld MT. Cost-effectivenessanalysis in the assessment of diagnostic imaging technologies. Radiology2005;235:361-70.

3. Hollingworth W. Radiology cost and outcome studies: standard practiceand emerging methods. AJR Am J Roentgenol 2005;185:833-9.

4. Lim ME, O’Reilly D, Tarride JE, et al. Health technology assessment forradiologists: basic principles and evaluation framework. J Am Coll Radiol2009;6:299-306.

5. Tynan A, Berenson RA, Christianson JB. Health plans target advancedimaging services: cost, quality, and safety concerns prompt renewed over-sight. Center Studying Health Syst Change Iss Brief 2008;118:1-3.

6. Patti JA. Cost control is one of radiology’s greatest threats. J Am Coll

Radiol 2007;4:440-2.