Research Article Patterns of Cancer Genetic Testing: A Randomized Survey of Oregon Clinicians

Transcription

1 Hindwi Publishing Corportion Journl of Cncer Epidemiology Volume 2012, Article ID , 11 pges doi: /2012/ Reserch Article Ptterns of Cncer Genetic Testing: A Rndomized Survey of Oregon Clinicins Summer L. Cox, 1 Amy I. Zlot, 1 Kerry Silvey, 1, 2 Debi Elliott, 3 Tr Horn, 3 Amber Johnson, 3 nd Richrd F. Lemn 4 1 Oregon Genetics Progrm, Public Helth Division, Oregon Helth Authority, Portlnd, OR 97232, USA 2 Oregon Center for Children nd Youth with Specil Helth Needs, Child Development nd Rehbilittion Center, Oregon Helth & Science University, Portlnd, OR 97239, USA 3 Survey Reserch Lb, Office of Reserch nd Sponsored Projects, Portlnd Stte University, Portlnd, OR 97201, USA 4 Center for Public Helth Prctice, Public Helth Division, Oregon Helth Authority, Portlnd, OR 97232, USA Correspondence should be ddressed to Summer L. Cox, summer.l.cox@stte.or.us Received 16 Mrch 2012; Revised 11 June 2012; Accepted 12 June 2012 Acdemic Editor: Suznne C. O Neill Copyright 2012 Summer L. Cox et l. This is n open ccess rticle distributed under the Cretive Commons Attribution License, which permits unrestricted use, distribution, nd reproduction in ny medium, provided the originl work is properly cited. Introduction. Approprite use of genetic tests for popultion-bsed cncer screening, dignosis of inherited cncers, nd guidnce of cncer tretment cn improve helth outcomes. We investigted clinicins use nd knowledge of eight brest, ovrin, nd colorectl cncer genetic tests. Methods. We conducted rndomized survey of 2,191 Oregon providers, sking bout their experience with fecl DNA, OncoVue, BRCA, MMR, CYP2D6, tumor gene expression profiling, UGT1A1, nd KRAS. Results. Clinicins reported low confidence in their knowledge of medicl genetics; most confident were OB-GYNs nd specilists. Clinicins were more likely to hve ordered/recommended BRCA nd MMR thn the other tests, nd OB-GYNs were twice s likely to hve ordered/recommended BRCA testing thn primry cre providers. Less thn 10% of providers ordered/recommended OncoVue, fecl DNA, CYP2D6, or UGT1A1; less thn 30% ordered/recommended tumor gene expression profiles or KRAS. The most common reson for not ordering/recommending these tests ws lck of fmilirity. Conclusions. Use of pproprite, evidencebsed testing cn help reduce incidence nd mortlity of certin cncers, but these tests need to be better integrted into clinicl prctice. Continued evlution of emerging technologies, dissemintion of findings, nd n increse in provider confidence nd knowledge re necessry to chieve this end. 1. Introduction Genomic medicine hs entered the clinicl setting. Currently vilble genomic 1 nd genetic tests enble disese surveillnce nd individully tilored tretment, nd mny more such tests re on the horizon. Chronic diseses, including brest, ovrin, nd colorectl cncer, hve multifctoril etiologies, including genetic components. In 2010, brest nd colorectl cncer were mong the four most commonly dignosed cncers nd were the second nd third most common cuses of cncer deth in both the USA nd in Oregon [1]. An estimted 5% 10% of ll brest nd ovrin cncers re hereditry, mening single gene muttion contributed to development of the cncer. The mjority of these inherited cncer cses re due to muttions in brest cncer susceptibility genes, which include BRCA1 nd BRCA2 (BRCA) [2]. Women within the generl popultion hve 12% lifetime risk of developing brest cncer nd 1% lifetime risk of developing ovrin cncer [3]. For women with BRCA muttions, however, the lifetime cncer risk is greter. It is estimted tht 47% 66% of women with BRCA1 muttions will develop brest cncer by ge 70, while 35% 46% of them will develop ovrin cncer by tht ge [4]. Risk of developing certin other cncers (e.g., pncretic cncer) lso increses mrkedly. Currently, identified muttions ccount for 5%-6% of colorectl cncer cses [5]. The generl popultion hs 6% lifetime risk of developing colorectl cncer, but for those with mismtch

2 2 Journl of Cncer Epidemiology repir gene (MMR) muttions, the risk increses to 80%, nd the risk of developing certin other cncers (e.g., endometril cncer) lso increses substntilly [6]. Morbidity from these heritble muttions plces substntil burden on both those who hve them nd on the helth cre system. Genetic tests tht cn be used for popultion-bsed cncer screening, dignosis of inherited cncer syndromes, nd selection of specific cncer therpies most likely to be effective for given ptient re now commercilly vilble [7 15]. In order to tke full dvntge of vlid nd cliniclly useful genetic tests, helth cre providers must not only become wre of them, but lso become knowledgeble bout their use nd interprettion. Providers must continue to improve their skills in ssessing fmily history nd other relevnt fctors to strtify ptient risk for specific cncers, improve decision mking for referrl to genetic specilists 2, nd decide when considertion of genetic testing is pproprite for given ptient [16 23]. The incresed use of direct-to-consumer mrketing for cncer-relted genetic tests mkes this doubly importnt, s clinicins re incresingly clled upon to interpret the results of genetic test tht my hve been ordered directly by their ptient rther thn helth cre provider. In the current helth cre milieu, providers in mny different settings cn order or recommend genetic test. Physicins or midlevel providers such s physicin ssistnts nd nurse prctitioners cn order these tests in primry cre setting, lternte or complementry cre providers such s nturopths my order them, nd specilists who primrily see ptients with cncer my order these tests s well. The fmilirity of these different provider groups with such tests, their ptterns in ordering them, nd their confidence in interpreting them my differ. Our survey evluted the extent to which helth cre providers in different prctice settings use eight commercilly vilble genetic tests to ssess personl or fmilil risk for brest, ovrin, nd colorectl cncer, or to guide tretment for these conditions. We lso explored providers rtionle for ordering/recommending these tests, their resons for not ordering/recommending these tests if they refrin from doing so, their level of confidence in their knowledge of medicl genetics, nd whether they refer to genetic specilists. The genetic tests we evluted fll into three ctegories: (1) popultion-bsed cncer screening, (2) refined risk ssessment for specific cncers in ptients lredy identified s high risk due to fmily or personl history, nd (3) testing to guide cncer tretment decisions. Inourstudy,wesurveyedprovideruseoffeclDNA, OncoVue, BRCA, MMR, CYP2D6, brest cncer tumor gene expression profiling, UGT1A1, nd KRAS testing. Tble 1 lists ech of the tests, describes them, nd summrizes the evidence-bsed recommendtions published by the Evlution of Genomic Applictions in Prctice nd Prevention (EGAPP), the U.S. Preventive Services Tsk Force (USP- STF), nd the Ntionl Comprehensive Cncer Network (NCCN). The evidence regrding the clinicl utility nd potentil hrms ssocited with these tests continues to grow. Guidelines for these tests rnge from recommending use of the test in specific circumstnces, to concluding there is insufficient evidence to recommend for or ginst using the test, to recommending ginst use of the test. The recommendtions conflict for some tests. It is importnt for clinicins to understnd both the benefits nd limittions of testing nd be wre of the importnce of pre- nd posttest counseling. A better understnding of how, nd why, helth cre providers use cncer genetic tests cn inform policy development nd eductionl efforts to ensure the pproprite nd effectiveuse of thesetests. 2. Methods 2.1. Study Popultion. The Oregon Genetics Progrm conducted the 2010 Oregon Helth Cre Provider Survey in collbortion with the Portlnd Stte University Survey Reserch Lb. We generted strtified rndom smple of primry cre providers nd specilists prcticing in Oregon to evlute the use of eight genetic tests for brest, ovrin, nd colorectl cncer. We used the 2010 licensee dtbses from the Oregon Medicl Bord, the Oregon Bord of Nturopthic Exminers, nd the Oregon Stte Bord of Nursing to identify possible respondents. Becuse the bords vry in their levels of specificity for prctice specilty, in order to survey subspecilists who tret cncer, for exmple, brest surgeons, we sent surveys to some providers who re unlikely to screen or tret for brest, ovrin, or colorectl cncer, for exmple, hed nd neck surgeons. To trget our study to providers who screen or tret for brest, ovrin, nd colorectl cncer, we sked clinicins whether they recommended screening or treted for brest, ovrin, or colorectl cncer in both screening postcrd nd on the survey. We excluded ny respondent who reported neither recommending screening nor treting brest, ovrin, or colorectl cncer. We lso sked clinicins to self-identify their specilty on the survey nd removed surveys from the nlysis if the responding clinicins indicted tht they did not belong to one of the helth cre provider groups of interest. We strtified the potentil respondents into four provider groups: primry cre providers, nturopths 3 obstetricins/gynecologists (OB-GYNs), nd specilists. Primry Cre Providers consisted of fmily physicins, internl medicine physicins, primry cre (generl prctice, fmily medicine, or fmily prctice) nurse prctitioners, nd primry cre physicin ssistnts. We nlyzed Nturopths seprtely to ssess ptterns of cre mong this growing clss of lternte/complementry cre providers. While one might consider obstetrics/gynecology primry cre specilty, we developed seprte strtum for this group becuse of the frequency with which they evlute ptients for possible ovrin or brest cncer. Specilists consisted of surgeons, colorectl surgeons, generl surgeons, gstroenterologists, oncologists, nd gynecologic oncologists Survey Instrument. A questionnire ws developed, piloted, nd miled to 2,506 Oregon helth cre providers in We sent up to three milings to recipients. The first miling included prenotifiction letter with endorsements

3 Journl of Cncer Epidemiology 3 Tble 1: Ntionl guidelines for cncer genetic tests included in the 2010 Oregon Helth Cre Provider Survey. Test Description Recommendtion Popultion-bsed screening for specific cncers OncoVue Tests for single nucleotide polymorphisms ssocited with incresed brest cncer risk. No recommendtions from EGAPP, NCCN, oruspstf. Test designed to screen for colorectl cncer, hs better (i) NCCN considers use of fecl (stool) DNA testing to be n option, but does not recommend it s first-line Fecl DNA sensitivity thn the trditionl fecl occult blood test screening tool [7]. (FOBT), nd my be more cceptble to the public thn (ii) USPSTF found insufficient evidence to recommend colonoscopy. use of fecl DNA testing s screening method for colorectl cncer [14]. Further ssessing risk for developing specific cncers in previously identified high-risk popultions BRCA MMR BRCA Tumor gene expression profiles CYP2D6 MMR UGT1A1 KRAS Tests designed to detect specific BRCA muttions ssocited with incresed risk for brest nd ovrin cncers. Providers use results to guide brest nd relted cncer prevention efforts. (i) NCCN nd USPSTF recommend BRCA testing for ptients t incresed risk of developing brest nd/or ovrin cncer due to fmily history [8, 14]. (i) EGAPP recommends genetic testing for Lynch syndrome in individuls with newly dignosed colorectl Testing for Lynch syndrome (previously known s cncer to reduce morbidity nd mortlity in reltives. HNPCC) includes testing of one or ll of the most They found insufficient evidence to recommend specific common mismtch repir genes (MMR) MCH1, MSH2, testing strtegy [11]. MSH6, nd PMS2. Providers use results to guide cncer (ii) NCCN recommends testing for Lynch syndrome for prevention efforts. individuls who meet certin criteri. The testing strtegy will depend on whether there is known MMR muttion in the fmily [7]. Guiding cncer tretment decisions in those lredy dignosed with cncer Test designed to detect specific BRCA muttions ssocited with incresed risk of ggressive, recurrent cncers. Providers use results to guide tretment decisions for people with brest, ovrin, nd relted cncers. Three tests, Oncotype DX, MmmPrint, nd H/I rtio, re currently being mrketed to help women with brest cncer nd their providers mke tretment decisions nd estimte risk of cncer recurrence. Test designed to help determine whether tmoxifen is likely to be useful therpy in those with estrogen receptor-positive brest cncer. Testing for Lynch Syndrome (previously known s HNPCC) includes testing of one or ll of the most common mismtch repir genes (MMR) MCH1, MSH2, MSH6 nd PMS2. Providers use the results to guide cncer mngement efforts. Test designed to help identify colorectl cncer ptients who re t incresed risk for n dverse rection to irinotecn therpy nd llow for chnges in mngement (e.g., drug choice, dosge). Testing for KRAS gene muttions my help identify colorectl cncer ptients who my not respond well to EGFR-inhibiting drugs such s pnitumumb (Vectibix) nd cetuximb (Erbitux). (i) NCCN recommends BRCA testing when the ptient meets certin personl nd fmily brest nd/or ovrin cncer history criteri [8]. (i) EGAPP found insufficient evidence to dvise for or ginst the use of tumor gene expression profiles in women with brest cncer [12]. (i) No recommendtions from EGAPP, NCCN, or USPSTF. (i) EGAPP recommends genetic testing for Lynch syndrome in individuls with newly dignosed colorectl cncer. They found insufficient evidence to recommend specific testing strtegy [11]. (ii) NCCN recommends testing for Lynch syndrome for individuls who meet certin criteri h. (i) EGAPP found insufficient evidence to recommend use of UGT1A1 genotyping in ptients with metsttic colorectl cncer treted with irinotecn [10]. (i) NCCN recommends testing for KRAS tumor gene sttus in ptients with metsttic colorectl cncer before inititing tretment with pnitumumb or cetuximb [24]. from leders from ech prctice group surveyed nd postge-pid screening postcrd. The postcrd sked if the clinicin recommended screening or treted brest, ovrin, or colorectl cncer; we removed respondents who returned the postcrd s ineligible from the miling survey. The survey ws sent in the second miling vi priority mil with $10 csh incentive, postge-pid return envelope, cover letter including n electronic link of the survey (so clinicins could respond either electroniclly or in written form), nd the endorsement letter. We did not send this second miling to respondents who hd lredy responded online; we sent those respondents check for $10 seprtely.

4 4 Journl of Cncer Epidemiology In the third miling, we sent postcrd to nonresponders with link to the web-bsed survey. We mde follow-up phone clls to nonresponders from both the first nd the second round of survey milings. In the survey, we sked questions bout provider demogrphics, length of time in prctice, prctice setting, the provider s level of confidence in their knowledge of medicl genetics, referrl to genetic specilists, nd use of fmily history to ssess risk for brest, ovrin, nd colorectl cncer. Portlnd Stte University Humn Subjects Reserch Review Committee pproved the survey, the informed consent process nd the dt collection protocol Survey Mesures. We determined whether to sk clinicin bout prticulr test by the clinicin s nswers to four questions. (1) Providers who nswered yes to the question, In your prctice, do you recommend brest nd/or ovrin cncer SCREENING to ptients without cncer? were sked bout OncoVue nd BRCA testing. (2) Providers who nswered yes to the question, In your prctice, do you recommend colorectl cncer SCREENING to ptients without cncer? were sked bout fecl DNA nd Lynch Syndrome genetic testing. (3) Providers who nswered yes to Do you TREAT ptients for brest nd/or ovrin cncer? were sked bout BRCA, brest cncer tumor gene expression profile, nd CYP2D6 testing. (4) Providers who nswered yes to Do you TREAT ptients for colorectl cncer? were sked bout MMR, UGT1A1, nd KRAS testing. Though we sked respondents whether they tret ptients for cncer, these questions did not specificlly define the nture of the tretment rendered. Therefore, these questions could hve been interpreted to include ncillry cre for pin, mngement of sequele of chemotherpy or surgery, or other types of cre. When exmining whether clinicin ordered specific test, we defined ordering test s ctully plcing n order to hve the test performed. We included the term recommending, llowing for the circumstnces where (1) the provider who discusses the test with the ptient is different from the provider tht ctully orders the test, or (2) the test is not conducted, but the provider recommended the test be done. Among providers who reported they recommended screening for brest, ovrin, or colorectl cncer, but did not order or recommend the corresponding tests, we exmined their rtionle for not ordering or recommending OncoVue nd fecl DNA. For those who reported treting brest, ovrin, or colorectl cncer, but did not order the corresponding tests, we exmined their rtionle for not ordering or recommending brest cncer tumor gene expression profiles, CYP2D6, UGT1A1, nd KRAS. Brief explntions were provided for some tests in the survey, for exmple, Hve you ever ordered or recommended n OncoVue test (e.g., multigene screening pnel for ptients without brest cncer) to determine ptient s brest cncer risk? We lso sked bout ech provider s rtionle for or ginst ordering or recommending ech of the tests. On questions regrding the rtionle for ordering or recommending BRCA or MMR testing, we clssified lwys usully, or sometimes responses s yes, nd never responses s no. We clssified respondents s referring to genetic specilist for BRCA or MMR testing if they responded tht they lwys or usully referred to genetic specilist. We did not sk resons for not ordering or recommending BRCA nd MMR testing on the survey Potentil Covrites. We sked bout potentil covrites which my ffect ssocitions between genetic testing nd provider group. Demogrphic covrites include vribles such s helth cre providers ge, sex, yers since forml trining, nd whether they recommend screening or tret for brest, ovrin, nd colorectl cncer. Prctice covrites include vribles such s number of ptients seen per week, prctice environment, nd geogrphic loction of helth cre clinic Dt Anlysis. We compred respondent self-reported prctice specilty nd credentils to their ssigned provider group (primry cre providers, nturopths, OB-GYNs, nd specilists), which ws bsed on their specilty, designted by the Oregon licensing bords. We moved surveys of three respondents from temporry other ctegory into the provider group tht better reflected their prctice specilty nd credentils. We excluded thirty-three surveys from further nlysis becuse the respondents indicted tht they prcticed in one of the provider groups tht typiclly do not screen or tret for brest, ovrin, or colorectl cncer (e.g., emergency medicine or nesthesiology). We clssified respondents s ordering or recommending BRCA nd MMR tests if they reported ordering or recommending the test t lest once in the 12 months prior to completing the survey. We clssified respondents who hd ever ordered or recommended OncoVue, fecl DNA, brest cncer tumor gene expression profile, CYP2D6, UGT1A1, nd KRAS tests s ordering or recommending these tests. We used Person χ 2 tests nd logistic regression to ssess the ssocition between provider group nd ordering or recommending cncer genetic tests, in ddition to resons why they chose to order/recommend or not order/recommend these tests. We used logistic regression to clculte djusted odds rtios (AOR) tht compred the odds of ordering or recommending genetic testing by the provider group, using primry cre providers s the referent ctegory. Covrites were included in these models if they were significntly ssocited with the provider group nd ordering or recommending genetic tests. We kept only covrites tht chnged the point estimte of the AOR by t lest 10% (compred with the full model) in the finl models. We did not present ssocitions between the covrites nd ordering genetic testing in this pper, s we were specificlly interested in the reltionship between the provider group nd genetic testing. All nlyses were performed using Stt version 19.0 [25]. We reported smple sizes (number of survey respondents) nd percentges s unweighted numbers nd estimtes becuse the smpling methodology eliminted the need for weighting.

5 Journl of Cncer Epidemiology 5 3. Results Of the 2,191 helth cre providers who received the survey, 1,242 returned the survey fully or prtilly completed, giving us response rte of nerly 57%, grtifying response for helth cre provider survey with the modest incentive of $10. We defined both pper nd web surveys s being fully completed if 80% 100% of pplicble questions were nswered nd prtilly completed if 50% 79% of pplicble questions were nswered. Though prtilly completed surveys were used, fully completed surveys mde up more thn 95% of the returned surveys. After the exclusions described in the Section 2, the finl smple included 1,209 respondents. Response rtes were similr mong ll provider groups. Tble 2() shows selected demogrphic nd prctice chrcteristics of our respondents by provider group. Among those who recommended brest, ovrin, or colorectl cncer screening, specilists were much more likely to report recommending screening ptients for colorectl cncer compred to brest nd ovrin cncer. Similrly, specilists were more likely to report treting ptients for colorectl cncer compred to brest nd ovrin cncer. About onethird of nturopths reported tht they tret ptients for brest, ovrin, nd/or colorectl cncer, 13% of primry cre providers nd 16% of OB-GYNs reported tht they tret ptients for brest cncer, nd 13% of primry cre providers reported tht they tret ptients for colorectl cncer. Tble 2(b) outlines providers confidence in their knowledge of medicl genetics by provider group. OB-GYNs hd the highest level of confidence in their knowledge of brest nd ovrin cncer genetics nd specilists hd the highest level of confidence in their knowledge of colorectl cncer genetics. Tble 2(c) shows the respondent referrl to genetic specilist when they suspect BRCA or MMR muttion by provider group. OB-GYNs nd specilists hd higher proportions who reported referring to genetic specilists for BRCA or MMR testing compred to nturopths nd primry cre providers. Among helth cre providers who report they recommend screening for brest nd ovrin cncer, lmost 3% reported they hd ordered or recommended OncoVue t lest once, nd mong clinicins who report recommending screening for colorectl cncer, 4% hd, t lest once, ordered or recommended fecl DNA screening. Among helth cre providers who tret brest nd ovrin cncer, 28% hd ordered or recommended brest cncer tumor gene expression profile test, while nerly 9% hd ordered or recommended CYP2D6 testing. Among clinicins who tret colorectl cncer, 20% nd lmost 4% hd ordered or recommended KRAS nd UGT1A1 testing respectively. Tble 3 outlines clinicin likelihood to report ordering or recommending BRCA or MMR tests in the pst 12 months, by provider group. OB-GYNs were more thn twice s likely to order or recommend BRCA testing in the 12 months prior to completing the survey for ptients without brest nd ovrin cncer thn primry cre providers. There were no sttisticlly significnt differences between provider groups in ptterns of ordering or recommending MMR testing for ny ptients or BRCA testing for ptients with cncer. The covrites tht were included in the finl models re described in the footnotes of Tble 3. The reson most often reported for not ordering n OncoVue, fecl DNA, brest cncer tumor gene expression profile, CYP2D6, UGT1A1,orKRAS test ws lck of fmilirity with the genetic test. About 10% of helth cre providers reported tht cost or insurnce noncoverge ws reson for not ordering or recommending OncoVue nd fecl DNA testing. In ddition, 17% nd 20% of providers reported tht prctice guidelines did not include OncoVue nd fecl DNA testing, respectively. Over one-third of helth cre providers reported not ordering or recommending CYP2D6, UGT1A1, nd KRAS testing becuse these tests were not relevnt to their ptients (Tble 4). A mjority of clinicins who reported ordering or recommending BRCA testing did so for the following resons: the ptient met prctice guidelines (82% 86%), to guide future screening decisions (75% 80%), to guide prophylctic mngement decisions (76% 80%), nd becuse their ptient requested the test (79% 81%). Clinicins gve the sme resons for ordering or recommending MMR testing, lthough the frequencies for ech reson were lower thn for BRCA testing (between 40% 73%) (Tble 5). We chose not to report resons for ordering or recommending OncoVue nd fecl DNA tests (tht could be used in popultion-bsed screening for specific cncers) or brest cncer tumor gene expression profiles, CYP2D6, UGT1A1, nd KRAS tests (tht could be used to guide cncer tretment decisions), becuse the smples were too smll to be relible. 4. Discussion There is pucity of peer-reviewed studies ssessing the clinicl knowledge nd use of the eight tests we investigted. Of ll of the tests, BRCA hs been the most studied, yet it remins underutilized. Indeed, our study suggests the likely underuse of certin tests (e.g., BRCA nd MMR), which re recommended for risk strtifiction in people t high risk for brest, ovrin, nd colorectl cncers. It lso highlights importnt brriers to pproprite testing, such s lck of confidence in genetics knowledge nd lck of fmilirity with recommended genetic tests (e.g., KRAS testing when deciding whether to tret ptient with cetuximb). Additionlly, our study suggests the ppropritely low use of tests where there re either guidelines recommending ginst use, guidelines stting tht there is insufficient informtion to recommend for or ginst use, no guidelines, or conflicting guidelines (e.g., OncoVue, fecl DNA, brest cncer tumor gene expression profiles, CYP2D6,nd UGTA1). The most common reson offered by clinicins for ordering or recommending BRCA nd MMR ws tht the ptient met prctice guidelines, indicting tht mny providers re wre of ntionl recommendtions regrding genetic testing nd consider these recommendtions in mking decisions bout testing. Still, in settings where testing would be recommended by multiple ntionl orgniztions, sizble portion of clinicins mke no reference to prctice guidelines s bsis for ordering or recommending BRCA or MMR testing, suggesting tht substntil gps in wreness remin.

6 6 Journl of Cncer Epidemiology Tble 2: Summry dt from the 2010 Oregon Helth Cre Provider Survey, by provider group. PCPs c () Demogrphic nd prctice chrcteristics,b. Nturopths OB-GYNs Specilists f Totl 363 (30.0%) 216 (17.9%) 333 (27.5%) 297 (24.6%) (27.5% 32.7%) (15.8% 20.1%) (25.1% 30.1%) (22.2% 27.1%) Men ge (yers) 357 (48.0 yrs) 211 (43.5 yrs) 329 (47.8 yrs) 288 (47.6 yrs) (26yrs 76yrs) g (28yrs 70yrs) g (27yrs 80yrs) g (27yrs 79yrs) g Number of ptients seen per week < >75 Recommend BOC h screening to ptients w/o cncer Recommend CRC i screening to ptients w/o cncer 104 (28.8%) 185 (86.4%) 103 (31.0%) 127 (43.3%) (24.4% 33.8%) (81.2% 90.4%) (26.3% 36.2%) (37.8% 49.1%) 129 (35.7%) 25 (11.7%) 126 (38.0%) 121 (41.3%) (31.0% 40.9%) (8.0% 16.7%) (32.9% 43.3%) (35.8% 47.0%) 127 (35.5%) 4(1.9%) 103 (31.0%) 45 (15.4%) (30.5% 40.4%) (0.7% 4.9%) (26.3% 36.2%) (11.7% 20.0%) 394 (97.5%) 196 (92.0%) 326 (98.5%) 187 (63.6%) (95.2% 98.7%) (87.5% 95.0%) (96.4% 99.4%) (57.9% 68.9%) 355 (98.3%) 204 (95.8%) 316 (94.9%) 275 (93.5%) (96.4% 99.3%) (92.1% 97.8%) (91.9% 96.8%) (90.1% 95.8%) Tret ptients for BOC h 47 (13.1%) 79 (36.9%) 53 (16.1%) 172 (58.3%) (10.0% 17.0%) (30.7% 43.6%) (12.5% 20.4%) (52.6% 63.8%) Tret ptients for CRC i 48 (13.3%) 63 (29.4%) (81.1%) (10.1% 17.2%) (23.7% 35.9%) (76.3% 85.2%) Ctegory totls my be less thn the totl number of respondents, due to missing vlues. b Bolded estimtes indicte significnt findings. c PCPs: primry cre providers which include fmily physicins, internl medicine physicins, primry cre nurse prctitioners, nd primry cre physicin ssistnts. d The column % reflects the percent responding within ech prctice ctegory. e CI: confidence intervl. f Specilists consisted of surgeons, colorectl surgeons, generl surgeons, gstroenterologists, oncologists, nd gynecologic oncologists. g Rnge in yers. h BOC: brest nd ovrin cncer. i CRC: colorectl cncer. PCPs c Confidence in personl knowledge of BOC g genetics Nottll Somewht Modertely/very (b) Confidence in personl knowledge of medicl genetics,b. Nturopths OB-GYNs Specilists f 111 (30.6%) 92 (42.6%) 23 (6.9%) 87 (30.0%) (26.1% 35.5%) (36.2% 49.3%) (4.6% 10.2%) (25.0% 35.5%) 188 (51.8%) 90 (41.7%) 132 (39.6%) 115 (39.7%) (46.6% 56.9%) (35.3% 48.4%) (34.5% 45.0%) (34.2% 45.4%) 64 (17.6%) 34 (15.7%) 178 (53.5%) 88 (30.3%) (14.1% 21.9%) (11.5% 21.2%) (48.1% 58.8%) (25.3% 36.0%)

7 Journl of Cncer Epidemiology 7 PCPs c Confidence in personl knowledge of CRC h genetics Not t ll Somewht Modertely/very (b) Continued. Nturopths OB-GYNs Specilists f 110 (30.5%) 114 (52.8%) 77 (23.2%) 43 (11.3%) ( ) ( ) ( ) ( ) 190 (52.6%) 81 (37.5%) 176 (53.0%) 105 (36.0%) (47.5% 57.7%) (31.3% 44.2%) (47.6% 58.3%) (30.7% 41.6%) 61 (16.9%) 21 (9.7%) 79 (23.8%) 154 (52.7%) (13.4% 21.1%) ( ) ( ) ( ) Ctegory totls my be less thn the totl number of respondents, due to missing vlues. b Bolded estimtes indicte significnt findings. c PCPs: primry cre providers which include fmily physicins, internl medicine physicins, primry cre nurse prctitioners, nd primry cre physicin ssistnts. d The column % reflects the percent responding within ech prctice ctegory. e CI: confidence intervl. f Specilists consisted of surgeons, colorectl surgeons, generl surgeons, gstroenterologists, oncologists, nd gynecologic oncologists. g BOC: brest nd ovrin cncer. h CRC: colorectl cncer. Refer g ptientsw/ocncerto genetic specilist for BRCA testing Refer g ptients w/cncer to genetic specilist for BRCA testing Refer g ptientsw/ocncerto genetic specilist for MMR testing Refer g ptients w/cncer to genetic specilist for MMR testing (c) Referrl of ptients to genetic specilist,b. PCPs c Nturopths OB-GYNs Specilists f 111 (67.7%) 27 (50.0%) 225 (87.5%) 82 (75.9%) (60.1% 74.4%) (36.9% 63.1%) (82.9% 91.1%) (67.0% 83.1%) 19 (55.9%) 25 (53.2%) 40 (87.0%) 121 (75.6%) (39.1% 71.5%) (39.0% 66.9%) (73.8% 94.0%) (68.3% 81.7%) 30 (56.6%) NA h 105 (91.3%) 122 (77.2%) (43.0% 69.2%) (84.5% 95.3%) (70.0% 83.1%) NA h NA h 131 (81.9%) 138 (78.0%) (75.1% 87.1%) (71.2% 83.5%) Ctegory totls my be less thn the totl number of respondents, due to missing vlues. b Bolded estimtes indicte significnt findings. c PCPs: primry cre providers which include fmily physicins, internl medicine physicins, primry cre nurse prctitioners, nd primry cre physicin ssistnts. d The column % reflects the percent responding within ech prctice ctegory. e CI: confidence intervl. f Specilists consisted of surgeons, colorectl surgeons, generl surgeons, gstroenterologists, oncologists, nd gynecologic oncologists. g Among providers who suspect muttion, those who lwys or usully refer to genetic specilist. h Unble to report estimtes due to smll cell size. It is lso worth noting tht mny providers reported ordering or recommending BRCA nd MMR testing in response to requests from ptients. A 2003 study ssessing the impct of pilot direct-to-consumer mrketing cmpign for BRCA testing in Atlnt, Denver, Rleigh-Durhm, nd Settle found tht providers perceived n increse in ptient wreness of testing, noted n increse in ptient requests for testing, nd ordered more BRCA tests, but there ws no chnge in rte of referrl to genetic specilists [26, 27]. Such referrls llow ptients considered to be t high risk to receive guidnce from helth professionl well grounded in cncer genetics bout wht tests would be most pproprite, s well s pre- nd post-test counseling [8, 14, 28]. Providers who do not specilize in genetics would seem to be importnt sources of such referrls. However, we found tht one-third to one-hlf of primry cre nd nturopthic providers in our smple did not mke such referrls, even when they suspected ptients were t n incresed risk for serious hereditry cncer syndromes. Our finding tht providers re using severl genetic tests for which there re no prctice guidelines highlights the need for further evlution to determine the clinicl usefulness nd pproprite role of these genetic tests, including severl ddressed in our survey. There is some evidence tht gene expression profiling tests (e.g., Oncotype DX, MmmPrint, nd H/I rtio) my help estimte risk of recurrence nd guide

8 8 Journl of Cncer Epidemiology Tble 3: Likelihood tht clinicins reported ordering or recommending specific cncer genomic test in the pst 12 months, by provider group. Totl clinicins BRCA for ptients without BC g or OC h,i,j 176 (62.6%) BRCA for ptients with BC g or OC h,i,l 91 (63.6%) Lynch syndrome testing m for ptients 68 (49.6%) without cncer n,o Lynch syndrome testing m for ptients 56 (61.5%) with cncer n,p PCPs b n (column %) c djusted OR d 51 (53.7%) 1.0 (referent) 14 (66.7%) 1.0 (referent) 9 (25.0%) 1.0 (referent) 5 (45.5%) 1.0 (referent) Nturopths n (column %) c djusted OR d 14 (36.8%) 0.8 ( ) 14 (46.7%) 0.5 ( ) 4 (50.0%) 3.9 ( ) OB-GYNs n (column %) c djusted OR d 75 (77.3%) 2.1 ( ) 11 (64.7%) 0.5 ( ) 10 (50.0%) 2.6 ( ) Specilists f n (column %) c djusted OR d 36 (70.6%) 2.1 ( ) k 52 (69.3%) 0.7 ( ) 45 (61.6%) 2.2 ( ) NA q NA q 50 (65.8%) 1.5 ( ) Bolded estimtes indicte significnt findings. b PCPs: primry cre providers include fmily physicins, internl medicine physicins, primry cre nurse prctitioners, nd primry cre physicin ssistnts. c The column % reflects the percent responding within ech prctice ctegory. d OR: odds rtio. e CI: confidence intervl. f Specilists consisted of surgeons, colorectl surgeons, generl surgeons, gstroenterologists, oncologists, nd gynecologic oncologists. g BC: brest cncer. h OC: ovrin cncer. i Adjusted for number of ptients seen per week nd confidence in brest nd ovrin cncer genetics. j Among clinicins who recommend brest nd ovrin screening. k P>0.05. l Among clinicins who tret brest nd/or ovrin cncer. m Specificlly testing for mismtch repir (MMR) genes, which my include testing in MCH1, MSH2, MSH6, nd PMS2 genes. n Adjusted for confidence in knowledge of colorectl cncer genetics. o Among clinicins who recommend colorectl cncer screening. p Among clinicins who tret colorectl cncer. q Unble to report estimtes due to smll cell size. tretment decisions [12, 15, 28 30]. Testing for CYP2D6 nd UGT1A1 genotypes re intended to identify individuls with ltered functionlity in genes tht effect drug metbolism. Some uthors hve concluded tht these tests my be useful to helth cre providers in deciding which tretments to recommend [13, 15]. However, evidence for the clinicl utility of CYP2D6 nd UGT1A1 testing is not conclusive nd evidence-bsed ntionl guidelines hve not endorsed these tests [7, 10, 12, 28]. Iftestprovestobecosteffectivendtoledtoimproved clinicl outcomes, it must then be integrted into clinicl prctice if its potentil to reduce cncer morbidity nd mortlity is to be relized. As we hve seen in the cse of BRCA testing, the best known of the eight tests in our study nd included in ntionl guidelines for number of yers, such n inclusion is n importnt but not sufficient prt of this process. Other strtegies include endorsement by medicl societies, cretion of decision support tools, nd incorportion into current nd continuing medicl eduction [19, 21, 23, 31 33]. The lck of confidence by helth cre providers in their bsic knowledge of cncer genetics is noteworthy nd is consistent with other studies [17, 18, 21 23, 26]. Becuse of this, there is higher chnce tht tests will be ordered incorrectly or inppropritely nd my be misinterpreted by nongenetic specilist, which my significntly hmper proper risk mngement [34]. This suggests need for continued trining to give clinicins the necessry bckground to know when they should order given genetic test, how to correctly interpret the results, nd in wht situtions ptients should be referred to genetic specilist. The higher level of confidence in brest nd ovrin cncer genetics mong OB-GYNs is not surprising nd is consistent with reserch by Trivers et l., who found tht being n OB-GYN ws predictor of pproprite referrl to genetic specilists [35]. Others hve found tht specilists, such s oncologists nd OB-GYNs, re often more knowledgeble bout cncer genetics nd cncer risk ssessment thn primry cre providers [26, 27, 36 38]. Low levels of confidence in personl knowledge of cncer genetics, coupled with lower rtes of referrl to genetics specilists for high-risk ptients emphsize the need for further medicl genetics trining, especilly mong primry cre providers nd nturopths. There re severl limittions to this study. Firstly, given our cross-sectionl study design, we could not infer cuslity from our dt. Secondly, the survey nswers were selfreported nd therefore subject to recll bis. Thirdly, differences in the time intervl we used for different tests in sking clinicins whether they hd ordered or recommended the tests (i.e. ever or in the lst twelve months ) limit our bility to compre the use of ll eight genetic tests mongst ech other. Due to smll smple sizes, we were not lwys ble to present results for the genetic tests by provider group. Finlly, we did not collect informtion bout the nture of therpy offered by respondents who reported treting brest, ovrin, nd colorectl cncer. This mkes reported

9 Journl of Cncer Epidemiology 9 Tble 4: Resons why clinicins did not order or recommend vrious cncer genomic tests mong clinicins who do not order tht specific test. Not fmilir with test n (%) Clinicl outcomes would not chnge n (%) Costs too much/insurnce will not cover it n (%) Test not vlid n (%) Prctice guidelines do not include this test n (%) Test not relevnt to ptients n (%) OncoVue c,d 787 (79.0%) 24 (2.4%) 106 (10.6%) 11 (1.1%) 173 (17.4%) 75 (7.5%) (76.45% 81.4%) (1.6% 3.6%) (8.9% 12.7%) (0.6% 2.0%) (15.1% 19.9%) (6.1% 9.3%) Fecl DNA e 783 (71.8%) 37 (3.4%) 119 (10.9%) 47 (4.3%) 216 (19.8%) 57 (5.3%) (69.1% 74.4%) (2.5% 4.7%) (9.2% 12.9%) (3.3% 5.7%) (17.6% 22.3%) (4.1% 6.7%) Tumor gene 126 (50.6%) 14 (5.6%) 21 (8.4%) 6 (2.4%) 17 (6.8%) 21 (8.4%) expression profiles f,g (44.4% 56.8%) (3.3% 9.3%) (5.6% 12.6%) (1.1% 5.3%) (4.3% 10.7%) (5.5% 12.6%) CYP2D6 g 131 (42.1%) 10 (3.2%) 10 (3.2%) 4 (1.3%) 19 (6.1%) 96 (30.9%) (36.7% 47.7%) (1.7% 5.9%) (1.7% 5.9%) (0.5% 3.4%) (3.9% 9.4%) (26.0% 36.3) h UGT1A1 i 187 (58.4%) NA j 7 (2.2%) NA j 16 (5.0%) 126 (39.4%) (52.9% 63.7%) (1.0% 4.5%) (3.1% 8.0%) (34.1% 44.9%) k KRAS i 108 (40.3%) NA j 4 (1.5%) NA j 9 (3.4%) 127 (47.4%) (34.6% 46.3%) (0.6% 3.9%) (1.7% 6.4%) (41.4% 53.4%) l The column % reflects the percent responding within ech prctice ctegory. b CI: confidence intervl. c OncoVue is multigene screening pnel for ptients without brest cncer. d Among clinicins who recommend brest nd ovrin screening to ptients without brest cncer. e Among clinicins who recommend colorectl cncer screening to ptients without colorectl cncer. f Brest cncer tumor gene expression profiles include Oncotype DX, MmmPrint, nd H/I rtio. g Among clinicins who tret ptients for brest cncer. h Test not relevnt to ptients becuse clinicin does not prescribe tmoxifen to ptients. i Among clinicins who tret ptients for colorectl cncer. j Unble to report estimtes due to smll cell size. k Test not relevnt to ptients becuse clinicin does not prescribe irinotecn to ptients. l Test not relevnt to ptients becuse clinicin does not prescribe nti-egfr therpytoptients. Tble 5: Resons why clinicins reported ordering specific cncer genomics tests in the pst 12 months, mong clinicins who ordered the genetic tests. Ptient met prctice guidelines n (%) Guide future screening decisions n (%) Guide prophylctic tretment decisions n (%) Ptient specificlly requests it n (%) BRCA for ptients 207 (85.5%) 192 (79.7%) 192 (79.7%) 198 (81.1%) without BC c or OC d,e (80.5% 89.5%) (74.1% 84.3%) (74.1% 84.3%) (75.7% 85.6%) BRCA for ptients with 100 (82.0%) 91 (75.2%) 93 (76.2%) 97 (78.9%) BC c or OC d,f (74.0% 87.9%) (66.6% 82.2%) (67.8% 83.0%) (70.6% 85.3%) Lynch syndrome testing g 72 (63.7%) 69 (60.5%) 68 (58.6%) 65 (57.0%) for ptients w/o CRC h,i (54.3% 72.2%) (51.2% 69.2%) (49.3% 67.3%) (47.7% 65.9%) Lynch syndrome testing g 50 (72.5%) 46 (67.6%) 27 (40.3%) 32 (50.0%) for ptients with CRC h,j (60.5% 81.9) (55.4% 77.9%) (29.0% 52.7%) k (37.7% 62.3%) The column % reflects the percent responding within ech prctice ctegory. b CI: confidence intervl. c BC: brest cncer. d OC: ovrin cncer. e Among clinicins who recommend brest nd ovrin screening to ptients without brest cncer. f Among clinicins who tret brest nd ovrin cncer. g Specificlly testing for mismtch repir (MMR) genes, which my include testing in MCH1, MSH2, MSH6, nd PMS2 genes. h CRC: colorectl cncer. i Among clinicins who recommend colorectl cncer screening to ptients without colorectl cncer. j Among clinicins who tret colorectl cncer. k For Lynch syndrome testing for ptients with cncer, the phrsing ws guide chemotherpeutic tretment decisions.

10 10 Journl of Cncer Epidemiology differences in the frequency of treting such ptients difficult to interpret. Higher rtes of cncer tretment reported by nturopths compred with primry cre providers or OB- GYNs my involve the perception tht nturopthic efforts to improve the ptient s overll helth re component of cncer tretment, view tht my not hve been shred by most llopthic clinicins providing services other thn surgery or chemotherpy. 5. Conclusion Reducing morbidity nd mortlity due to brest, ovrin, nd colorectl cncers is ludble gol. Consistent use of evidence-bsed genetic tests could contribute to tht objective, while underutiliztion of these tests limits their potentil contribution. Perceived low levels of knowledge boutrelevntgeneticsppertobenobstclebothtothe use of these tests nd to the timely referrl to genetic specilists. Clinicins working in settings with higher volumes of cncer ptients note higher levels of confidence in relevnt knowledge of medicl genetics, but even then, lmost hlf report low confidence in their knowledge bse. Eduction through multiple modlities is resonble strtegy to ddress these perceived knowledge deficits. At this time, the pproprite role of severl genetic tests is undetermined, but for some genetic tests (e.g., BRCA nd MMR) the cost effectiveness, efficcy in guiding preventive cre nd tretment, nd beneficil helth outcomes hve been demonstrted. Continued evlution of emerging technologies nd subsequent dissemintion of informtion bout the clinicl utility, interprettion, nd indictions for the use of such tests re necessry to ensure their integrtion into pproprite ptient cre. Acknowledgments The Centers for Disese Control nd Prevention (CDC) Coopertive greement no. CDC-RFAGD (Grnt no. 1U38GD000061) Genomic Applictions in Prctice nd Prevention: Trnsltion Progrms in Eduction, Surveillnce, nd Policy supported this project. Endnotes 1. Generlly, genetic test ssesses for the presence or effect of single gene, while genomic test ssesses the presence or ctivity of multiple genes. In this pper, we will use the term genetic tests nd genetic testing to describe both genetic nd genomic tests. 2. The Americn College of Surgeons Commission on Cncer hs recently published new cncer progrm stndrds. These stndrds require tht cncer risk ssessment, genetic counseling, nd testing services be provided to ptients by qulified genetics professionl. The stndrds lso outline the criteri for pre- nd post-genetic counseling nd define the trining nd experience of qulified genetics professionls, whom we refer to s genetic specilists. The Cncer Progrm Stndrds 2012: Ensuring Ptient-Centered Cre is vilble t See especilly Stndrd 2.3, Risk Assessment nd Genetic Counseling on pg Nturopthic physicins use whole-body nd minimlly invsive pproch with the gol of restoring the helth of their ptients; their model of cre voids drugs nd surgery nd emphsizes the use of nturl gents nd physicl mens ( References [1] Americn Cncer SocietyCncer Fcts nd Figures, 2010, document/cspc pdf. [2] NCI Ntionl Cncer Institute, Genetics of Brest nd Ovrin Cncer PDQ Summry, 2012, professionl/. [3] Ntionl Cncer Institute (NCI), BRCA1 nd BRCA2: Cncer Risk nd Genetic Testing, 2009, [4] S. Chen nd G. Prmigini, Met-nlysis of BRCA1 nd BRCA2 penetrnce, Journl of Clinicl Oncology, vol. 25, no. 11, pp , [5] Ntionl Cncer Institute (NCI), Genetics of Colorectl Cncer PDQ Summry, 2012, [6] W. S. B. Kohlmnn, S. B. Gruber, nd L. Syndrome, GeneReviews t GeneTests: Medicl Genetics Informtion Resource, 2004, [7] Ntionl Comprehensive Cncer Network (NCCN), NCCN Clinicl Prctice Guidelines in Oncology Colorectl Cncer Screening, [8] Ntionl Comprehensive Cncer Network (NCCN), NCCN Clinicl Prctice Guidelines in Oncology Genetic/Fmilil High-Risk Assessment: Brest nd Ovrin, [9] Intergenetics Incorported, Wht is OncoVue, 2012, oncovue. [10] Evlution of Genomic Applictions in Prctice nd Prevention (EGAPP) Working Group, Recommendtions from the EGAPP working group: cn UCT1A1 genotyping reduce morbidity nd mortlity in ptients with metsttic colorectl cncer treted with irinotecn? Genetics in Medicine, vol. 11, no. 1, pp , [11] Evlution of Genomic Applictions in Prctice nd Prevention (EGAPP), Recommendtions from the EGAPP Working Group: genetic testing strtegies in newly dignosed individuls with colorectl cncer imed t reducing morbidity nd mortlity from Lynch syndrome in reltives, Genetics in Medicine, vol. 11, no. 1, pp , [12] Evlution of Genomic Applictions in Prctice nd Prevention (EGAPP), Recommendtions from the EGAPP Working Group: cn tumor gene expression profiling improve outcomes in ptients with brest cncer? Genetics in Medicine, vol. 11, no. 1, pp , 2009.

11 Journl of Cncer Epidemiology 11 [13] M. Tzvetkov nd N. von Ahsen, Phrmcogenetic screening for drug therpy: from single gene mrkers to decision mking in the next genertion sequencing er, Pthology, vol. 44, no. 2, pp , [14] U.S. Preventive Services Tsk Force (USPSTF), The Guide to Clinicl Preventive Services , [15] R. G. Wtson nd H. L. McLeod, Phrmcogenomic contribution to drug response, Cncer Journl, vol. 17, no. 2, pp , [16] D. B. White, V. L. Bonhm, J. Jenkins, N. Stevens, nd C. M. McBride, Too mny referrls of low-risk women for BRCA1/2 genetic services by fmily physicins, Cncer Epidemiology Biomrkers nd Prevention, vol. 17, no. 11, pp , [17] M. J. H. Brs, L. Hennemn, nd L. P. Ten Kte, Deficiency of knowledge of genetics nd genetic tests mong generl prctitioners, gynecologists, nd peditricins: globl problem, Genetics in Medicine, vol. 7, no. 9, pp , [18] N. Drury, J. Bethe, P. Guilbert, nd N. Qureshi, Genetics support to primry cre prctitioners demonstrtion project, Journl of Genetic Counseling, vol. 16, no. 5, pp , [19] B. S. Flynn, M. E. Wood, T. Ashikg, A. Stockdle, G. S. Dn, nd S. Nud, Primry cre physicins use of fmily history for cncer risk ssessment, BMC Fmily Prctice, vol. 11, rticle 45, [20] J. H. Medlie, S. J. Zyznski, M. A. Goodwin, nd K. C. Stnge, Two physicin styles of focusing on the fmily: their reltion to ptient outcomes nd process of cre, Journl of Fmily Prctice, vol. 49, no. 3, pp , [21] P. W. Rose, E. Wtson, P. Yudkin et l., Referrl of ptients with fmily history of brest/ovrin cncer GPs knowledge nd expecttions, Fmily Prctice, vol. 18, no. 5, pp , [22] R. D. Sifri, R. Wender, nd N. Pynter, Cncer risk ssessment from fmily history: gps in primry cre prctice, The Journl of Fmily Prctice, vol. 51, no. 10, p. 856, [23] S. Suther nd P. Goodson, Brriers to the provision of genetic services by primry cre physicins: systemtic review of the literture, Genetics in Medicine, vol. 5, no. 2, pp , [24] Ntionl Comprehensive Cncer Network (NCCN), NCCN Clinicl Prctice Guidelines in Oncology Colon Cncer, [25] SttCorp, Stt sttisticl softwre for professionls, Version 19.0, College Sttion, Tex, USA, [26] M. F. Myers, M. H. Chng, C. Jorgensen et l., Genetic testing for susceptibility to brest nd ovrin cncer: evluting the impct of direct-to-consumer mrketing cmpign on physicins knowledge nd prctices, Genetics in Medicine, vol. 8, no. 6, pp , [27] Centers for Disese Control nd Prevention (CDC) nd P.H.G., Genetic testing for brest nd ovrin cncer susceptibility: evluting direct-to-consumer mrketing Atlnt, Denver, Rleigh-Durhm, nd Settle, 2003, Morbidity nd Mortlity Weekly Report, pp , [28] Ntionl Comprehensive Cncer Network (NCCN), NCCN Clinicl Prctice Guidelines in Oncology Brest Cncer, [29] D. Fumglli, C. Desmedt, M. Igntidis et l., Gene profiling ssy nd ppliction: the predictive role in primry therpy, Journl of the Ntionl Cncer Institute Monogrphs, no. 43, pp , [30] L. J. vn t Veer, H. Di, M. J. Vn de Vijver et l., Gene expression profiling predicts clinicl outcome of brest cncer, Nture, vol. 415, no. 6871, pp , [31] L. Acheson, Fostering pplictions of genetics in primry cre: wht will it tke? Genetics in Medicine, vol. 5, no. 2, pp , [32] S. B. Hg, M. M. Crrig, J. M. O Dniel et l., Genomic risk profiling: ttitudes nd use in personl nd clinicl cre of primry cre physicins who offer risk profiling, Journl of Generl Internl Medicine, vol. 26, no. 8), pp , [33] S. E. Strus, J. Tetroe, nd I. Grhm, Defining knowledge trnsltion, Cndin Medicl Assocition Journl, vol. 181, no. 3-4, pp , [34] K. L. Brierley, D. Cmpfield, W. Ducine et l., Errors in delivery of cncer genetics services: implictions for prctice, Connecticut Medicine, vol. 74, no. 7, pp , [35] K. F. Trivers, L. M. Bldwin, J. W. Miller et l., Reported referrl for genetic counseling or BRCA 1/2 testing mong United Sttes physicins: Vignette-Bsed Study, Cncer, vol. 117, no. 23, pp , [36]T.Doksum,B.A.Bernhrdt,ndN.A.Holtzmn, Does knowledge bout the genetics of brest cncer differ between nongeneticist physicins who do or do not discuss or order BRCA testing? Genetics in Medicine, vol. 5, no. 2, pp , [37] A. N. Freedmn, L. Wideroff, L. Olson et l., US physicins ttitudes towrd genetic testing for cncer susceptibility, Americn Journl of Medicl Genetics, vol. 120, no. 1, pp , [38] L. Wideroff, S. T. Vdprmpil, M. H. Greene, S. Tplin, L. Olson, nd A. N. Freedmn, Hereditry brest/ovrin nd colorectl cncer genetics knowledge in ntionl smple of US physicins, Journl of Medicl Genetics, vol. 42, no. 10, pp , 2005.