Objective: To describe the impact on patients and physicians ata managed care organization (MCO) of a direct-to-consumeradvertising (DTC-ad) campaign concerning testing for the BRCA1and BRCA2 genes.
Study Design: Observational study.
Methods: In 2003, we mailed a 30-item questionnaire to 750randomly chosen female members of Kaiser Permanente Colorado(KPCO) aged 25 to 54 years, and 100 female KPCO members witha history of breast cancer genetic referral. We mailed a 7-item questionnaireto 180 randomly chosen KPCO primary care providers.
Results: Of 394 patient respondents, 245 (62%) reported exposureto the DTC-ad of whom 63% reported that the DTC-ad causedno anxiety at all. A high level of perceived breast cancer risk andbeing of Hispanic ethnicity each were independently associatedwith reported anxiety due to the DTC-ad (adjusted odds ratio[OR] = 3.23, 95% confidence interval [CI] = 1.35, 7.73, andadjusted OR = 4.19, 95% CI = 1.48, 11.83, respectively). Greaterknowledge was seen among respondents exposed to the DTC-adthan among those reporting no exposure (= .015). Of the physicianrespondents, 84% reported that the DTC-ad caused no strainon the doctor-patient relationship, and nearly 80% reported noeffect on daily clinical practice. Genetic referrals soared more than200% compared with the prior year, when there was no advertising.
Conclusion: The DTC-ad had a marked impact on genetic services,but little apparent negative impact on patients or primary careproviders at an MCO.
(Am J Manag Care. 2005;11:601-608)
In the mid-1990s germline mutations in the BRCA1and BRCA2 (BRCA1/2) genes were found to be associatedwith an increased risk for breast and ovariancancers, and Myriad Genetics Laboratories, Inc. (Myriad)obtained the patent for direct sequencing.1-3 Professionalsand lay advocates were greatly concerned about theethical ramifications of this patent, predicting that anysubsequent marketing campaign by the laboratorywould have mostly negative effects on women's healthcare.4 However, until 2002 there was no major marketingcampaign.5
In May 2002 Myriad began educational outreach toproviders in the Denver and Atlanta metropolitan areasto prepare them for a direct-to-consumer advertising(DTC-ad) campaign concerning testing for the BRCA1/2genes (BRACAnalysis), which would run from mid-September 2002 through mid-February 2003.5 These 2markets and the targeted demographic group, womenaged 25 to 54 years, were selected based on researchconducted by Myriad, which assessed factors rangingfrom Myriad's own infrastructure to the sophisticationof local genetic services.5 Myriad's intensive 5-monthcampaign used television, radio, and print media andwas projected to reach more than 90% of the targetedpopulation an average of 16.5 times each.
Advertisements presented 4 middle-aged femaleactors stating that breast cancer ran in their familiesand they had sought testing. Although technically correctin content and without distortion or exaggeration,the DTC-ad's explicit and implicit consumer testimonialsrepeatedly stressed that if breast cancer "runs" in awoman's family, she should undergo BRACAnalysis"now." The advertisements encouraged women to call atoll free number or talk with their doctor.
Controversy arose as Myriad initiated its campaign.6-10At issue: for every 10 000 women reached by the advertisements,only 15 would have a BRACA1/2 mutation thatconferred high risk.11 Although women of Jewish ancestryhave a higher risk of a mutation, the DTC-ad was targetedat the general population.5 With a low prevalence ofwomen in the general population having a BRCA1/2mutation, this advertising "overreach" carried potentialfor creating patient anxiety, contributing to patient misconception,and straining the patient-provider relationship.The literature suggesting limited knowledge ofpatients and providers about BRCA1/2 testing supportedthese concerns.12-14 However, such literature also can justifythe DTC-ad, because benefits of genetic servicesmay be high and such services are underutilized.15-21
In this study the effect of Myriad's inaugural marketingcampaign for BRACAnalysis on the patients andphysicians of Kaiser Permanente Colorado (KPCO) isassessed. Specific outcomes are patient anxiety andmisconception, strain on the doctor-patient relationship,and demand for healthcare services.
We received institutional review board approval for datacollection from KPCO, a nonprofit, closed-panel managedcare organization (MCO) serving nearly 370 000 membersin the Denver metropolitan area. Clinical referralguidelines for cancer genetic services have existed atKPCO since 1997 and require pretest genetic counseling.Referrals must be provider generated.
We mailed an anonymous survey in the spring of2003 to a random group of 750 female KPCO membersaged 25 to 54 years who had been members for at least33 months between February 2000 and January 2003.Because the general population is mostly not atincreased risk for a BRCA1/2 mutation, we oversampledfor higher risk by randomly selecting 100 additionalfemale members with a breast and ovarian cancergenetic referral made during the DTC-ad period.
The survey assessed exposure to the DTC-ad.Exposure was defined as having heard of BRACAnalysisor of genetic testing for hereditary breast or ovariancancer risk through a television, radio, newspaper, ormagazine advertisement. We identified 2 levels of exposure:any versus none.
Subjects reporting exposure were asked to rate on a3-point scale (very, somewhat, not at all) the emotionsthe advertisement made them feel. Based on literatureand focus group work by Myriad, these emotions includedanger, anxiety, sympathy, concern for self or family,confusion, skepticism, excitement, gratefulness forbeing told about genetic testing, and a sense of urgency.
Survey respondents categorized their perceived riskfor breast cancer as high, average, or low compared withthat of women their own age. Respondents also reportedon demographic factors (age, race/ethnicity, education,income, employment, marital status, numbers of siblingsand offspring, and level of spirituality), and healthcareutilization (use of routine care within previous 2 years,previous mammogram, previous breast biopsy).
For determination of actual BRCA1/2 mutation risk,we collected the risk factors of personal and family cancerhistories and Jewish ancestry, and used the Myriadmutation prevalence tables (MPTs).22 Although thereare other computer-based mutation probability assessmentmodels, the MPT (the only tool based on actual testresults) best fit the level of information available.22-24Using the MPT, we created a high-probability categoryfor patients equating to a 10% or greater chance of testingpositive for a BRCA1/2 mutation, a moderate categoryequating to a probability from 5% up to 10%, and a lowcategory for less than a 5% chance of testing positivewith BRACAnalysis.
We captured patient knowledge using 9 questionsfrom a previously validated scale of cancer geneticknowledge.25 Seven of the questions were specificallyaddressed in the DTC-ad content.
Also in 2003 we administered an anonymous writtensurvey to a random sample of physicians in the KPCOdepartments of obstetrics/gynecology, family practice,and general internal medicine (n = 180). In addition tocollecting medical specialty and number of years sincecompleting residency training, we asked physicians howmuch undue anxiety the DTC-ad created amongpatients (very, somewhat, none), how much time wasneeded to correct patients' misconceptions due to theDTC-ad, and how the DTC-ad affected the patient-providerrelationship. We asked how the DTC-ad affectedthe number of office visits and patient telephonecalls, and how those numbers compared with the numbersin the preceding year. We further asked how theDTC-ad affected routine clinical practice overall (verypositively, somewhat positively, no effect, somewhatnegatively, very negatively).
We assessed the increase in demand for geneticservices compared with the demand 1 year before theDTC-ad through numbers of provider-generated referrals.Referrals are tracked in an electronic geneticdatabase.
The chi-square test of proportions assessed the statisticalsignificance of association between perceivedrisk for breast cancer and patient emotions. For multivariateanalyses, we used logistic regression withreported emotion (any vs none) as the dependent variable and level of perceivedbreast cancer risk (high,average, low) as the independentvariable. Potentialconfounders in each modelincluded age, race/ethnicity,income, and education.Neither personalhistory of cancer nor levelof actual risk of a mutationsignificantly added tothe logistic regressionmodels. Other potentialdemographic confounderscaptured in the patientsurvey and potentialdemographic interactionsalso did not add to themodels. We found nocollinearity between thedemographic factors in thefinal models. Because thewomen with high risk for amutation were mostlysampled from a specialpopulation source, wemodeled the final associationsbetween perceivedrisk of breast cancer andemotions both with andwithout the high-risk subgroup.There was no significantchange in the results;thus, we kept the entirehigh-risk group in the finalmodels.
We created a patientknowledge score, definedas the number of correctanswers to the 9 knowledgequestions with missingand "Don't know" replies counting zero. Weassessed differences inknowledge scores by DTC-adexposure using theWilcoxon rank-sum test. We tested statistical associationbetween exposure and each of the knowledgequestions using the chi-square test.
For the physician survey, we used the chi-square testto determine differences among medical specialties inreported changes in clinical practice attributable to thead campaign.
The chi-square test was used to test for the significanceof the change in the number of genetic counselingreferrals attributable to DTC-ad. Specifically,the total number of referrals during the active advertisingperiod was compared with the total number inthe previous year, controlling for changes in averagemembership for women aged 25 to 54 years. All analyses were done using SAS version 8.02 (SAS Institute Inc,Cary, NC).
The patient survey received a 48% response ratewith 394 completed surveys and 25 undeliverable. Mostrespondents were in their 40s, and more than 80% werenon-Hispanic white (Table 1). The majority (78%) hadsome college or higher as the highest completed level ofeducation, and 91% had utilized healthcare serviceswithin the previous 2 years. We could only comparerespondents with nonrespondents by age, with 45 yearsas the median age for respondents and 43 years as themedian age for nonrespondents (data not shown).
Of the respondents, 245 (62%) reported exposure tothe DTC-ad. There were no significant demographic differencesbetween those exposed and not exposed tothe DTC-ad, with the exception that those reportingexposure were more likely to report greater previoushealthcare utilization (= .03; Table 1). There were nosignificant differences between respondents exposedand not exposed to the DTC-ad with respect to perceivedrisk of breast cancer or in actual BRCA1/2 mutationrisk (Table 1). The other demographic andutilization measures captured by the survey were notsignificantly associated with exposure (data not shown).
Among the women who reported exposure to theDTC-ad, 63% reported that the DTC-ad caused them noanxiety at all, 76% reported no confusion, and 71% reportedno urgency; however, 65% reported feeling somewhator very concerned. Each of these emotions was significantlyassociated with higher perceived risk for breastcancer (<.05 for all emotions; data not shown). Foreach of the other examined emotions, the overwhelmingresponse was of no effect, and there was no significantassociation with perceived risk (data not shown).
Logistic regression showed a high level of perceivedbreast cancer risk and Hispanic ethnicity each to beindependently associated with reported anxiety dueto the DTC-ad (adjusted odds ratio [OR] = 3.23, 95%confidence interval [CI] = 1.35, 7.73, and adjustedOR = 4.19, 95% CI = 1.48, 11.83, respectively) (Table2). Higher reported incomes were associated with lessreport of anxiety due to the DTC-ad. Logistic regressionshowed similar findings for the other reported emotions:concern was associated with high perceived riskfor breast cancer (OR = 6.09, 95% CI = 2.18, 16.98);confusion was associated with high perceived risk(OR = 3.71, 95% CI = 1.38, 9.98) and Hispanic ethnicity(OR = 2.57, 95% CI = 0.87, 7.54); and urgency wasassociated with high perceived risk (OR = 8.63, 95%CI = 3.37, 22.12) and Hispanic ethnicity (OR = 3.35,95% CI = 1.13, 9.97) (data not shown).
The median knowledge score of 6 did not differ byexposure. However, knowledge scores skewed higheramong the exposed group (5% and 95% quantiles of 3and 9 in the exposed group vs 1 and 8 in the unexposedgroup; = .015) (data not shown). The individual questionshad mixed association with DTC-ad exposure.Among the 7 questions specifically referred to in theactual advertisement, a large percentage of respondentsanswered "Don't know" regardless of exposure(Table 3).
Among the 180 physicians surveyed, we obtained 97completed surveys for a 54% response rate. The respondentswere mostly in internal medicine (40%), with 32%in family practice and 28% in gynecology (data notshown). More than 85% of the respondents were morethan 5 years out from completing residency training.
The majority of physician respondents (65%) reportedthat the DTC-ad caused no undue patient anxiety,which did not vary by medical specialty (= .96;Table 4). In regard to patient misconception, 58%reported that time spent to correct patient misconceptionswas not at all affected by the DTC-ad. By medicalspecialty, gynecologists were more likely to report needing"a little" more time to correct patient misconceptiondue to the DTC-ad (= .07; Table 4).
Whereas 80% of physicians reported that the DTC-addid not cause any strengthening of the doctor-patientrelationship, 84% reported that the DTC-ad caused nostrain (Table 4). Gynecologists were more likely toreport a strengthening of the relationship as a result ofthe DTC-ad (= .03), with no difference by specialty forreports of straining the relationship (= .80). Themajority of physician respondents (69%) reported thatthe DTC-ad caused no pressure to order a genetic referral,with no difference by medical specialty (= .94).
The majority (74%) of physicians reported no differencein number of patient office visits, and 81% reportedno difference in number of patient phone calls dueto the DTC-ad, with no difference by medical specialty(= .96 and .50, respectively; Table 4). Only 14% ofphysicians reported negative effects of the DTC-ad; 79%reported no effect on daily clinical practice.Gynecologists were more likely to report the DTC-adhad a positive effect on their clinical practice (= .04).
Genetic counseling referrals during the DTC-adincreased 240% compared with the same months 1 yearbefore the DTC-ad (data not shown). During theDTC-ad, fewer women with a high probability of amutation were referred than the year before (48% vs69%; <.01). The percentage of referred patientsundergoing counseling did not change between the 2time periods, and the percentage of testing uptake variedbut reflected the change in mutation probabilityamong referred patients.
Even after completion of Myriad's DTC-ad, controversyabout its potential effects continued.26,27 To ourknowledge, our study is the first to report effects basedon data, and overall, the data show that few negativeeffects actually occurred.
The DTC-ad caused little anxiety or confusion amongour patient survey respondents and little report ofpatient anxiety among the physician survey respondents.This may reflect Myriad's use of focus groups todesign an advertisement that did not frighten women.28
Although the DTC-ad caused little anxiety overall,those with higher risk perceptions reported more anxiety.This finding is consistent with breast cancer literatureon risk perception and anxiety, and may in partreflect the fact that the Myriad advertisement cast awide net when targeting women "with a family historyof breast cancer."29 Without a more specific definitionof family history, it is likely that women with higher perceivedbreast cancer risk would consider themselves tobe at a higher probability for a mutation, regardless oftheir true mutation risk, and feel some level of anxietywhen viewing ads for genetic testing.
Our finding that Hispanics, the fastest growing ethnicgroup in the United States, are more likely to have ananxious response to the advertisement merits furtherinvestigation. Although the DTC-ad was only in English,we do not believe a language barrier to be the most likelyexplanation, because our survey also was only inEnglish. The better potential explanation may be foundin cultural studies of cancer control interventions,which suggest that selected attitudes about canceramong Latinas fit a cultural theme of and thatthere is very little one can do to prevent getting cancer.30-32 Latinas also may be more concerned about thedisadvantages of genetic testing.33
Our patient knowledgefindings affirm previouswork showing misconceptionsabout BRCA1/2among the general population.A study conductedconcurrently by theCenters for Disease Controland Prevention on theeffects of Myriad's marketingcampaign reported littledifference in knowledgeof genetic testing betweenwomen in the general populationin the DTC-adcities versus women in thecontrol cities.34 Our studyassessed knowledge ofbreast cancer genetics (vstesting) by self-reportedexposure to the advertisements.Although thedifference in knowledgescore by exposure wasmodest, the DTC-ad mayhave contributed tolessening patient misconception.Similar todata regarding pharmaceuticalDTC advertising,the physicianfindings also supportthat this DTC-ad mayhave contributed to betterpatient understanding.35,36
More in-depth analysisof our work showed thatthe DTC-ad caused anincrease in demand forgenetic services at KPCO.37Compared with a healthplan in a non-DTC-admarket, KPCO saw amarked impact on cancergenetic services interms of overall volumeand on the risk level ofpatients being referred.37The impact on test ordering was less pronounced,likely due to the KPCO testing criteria and due toaccess delays, as the majority of the referral increasewas among lower risk women not meeting those testingcriteria.37
Our provider survey, however, additionally showsthat the DTC-ad led to little demand on primary careservices in that there was little reported increase inoffice visits or telephone calls during the active DTC-adperiod. This may in part reflect KPCO utilizing ascreener to identify family history that meets KPCOcancer genetic referral guidelines, thereby making iteasy for KPCO providers to make a referral.
While impact on daily volume did not increase, wealso found that the DTC-ad did not lead to moredemands on the physician in terms of pressure to ordera referral or strain on the patient-provider relationship.Further, we found no overall negative effect of the DTCadon physicians' daily practice. These data are differentin magnitude from pharmaceutical DTC-ad effectsand may indicate a difference between DTC-ad forgenetic versus nongenetic products.
Our survey work is limited by response rate, whichmay be attributable to lack of financial incentives. Withfew data on nonrespondents, we cannot be sure of theimpact of selection bias. For the patient survey, a $3video coupon was available only to survey respondents.Our physician survey was abbreviated to enhance theresponse rate, as we were unable to afford any incentive.The KPCO response may not be generalizable towomen in MCOs in other geographic regions or to thosewith health coverage where counseling and testing arenot covered benefits.
Perhaps more importantly, however, we did not capturethe effect of local media on patients' surveyresponses. Media in the Denver metropolitan areaincluded newspaper, television, and radio commentaries.We also did not capture awareness of benefit coverage.Such factors may have influenced women'sperceptions regarding their own risk, genetic testing,and the DTC-ad.
However, this study is the only effort to our knowledgethat quantifies the effect of the DTC-ad at thepatient, provider, and health-system levels. Further,MCOs are unique in their ability to assess these outcomesindependent of concerns about testing costs(more than $2700 in 2003) or about the financial implicationsof referral.
It appears that Myriad conducted a DTC-ad campaignwith little negative impact outside of targeting toobroad an audience, which drove patient demand forgenetic information. Future advertising, however, couldbe markedly different in content or framing and therebyhave markedly different outcomes. Thus, efforts fornational DTC-ad standards should continue, and theseefforts should consider rigorous study designs for moreconclusive understanding of benefits and harms.Additionally, providers and payers should be aware thatin the age of DTC-ad, provision of genetic services mayrequire new models of delivery.
The authors acknowledge Melissa L. Finucane, PhD, of Kaiser PermanenteCenter for Health Research, Honolulu, Hawaii, and Richard Meenan, PhD,MPH, of Kaiser Permanente Center for Health Research, Portland, Ore, fortheir content expertise in the development of the patient survey.
From the Kaiser Permanente Clinical Research Unit, Denver, Colo (JM, DPR, JE, AKR);the Department of Communications, Flagler College, St. Augustine, Fla (SL); and the HenryFord Health System, Detroit, Mich (SHA).
The design and conduct of this study were partially funded by contributions of theKaiser Permanente Colorado Clinical Research Unit and the Josephine F. Ford EndowedChair in Cancer Genetics. Data collection via the patient survey was partially funded byMyriad Genetics Laboratories, Inc.
Address correspondence to: Judy Mouchawar, MD, MSPH, Clinical ResearchUnit, Kaiser Permanente Colorado, 580 Mohawk Dr, Boulder, CO 80303. E-mail:firstname.lastname@example.org.
1. Miki Y, Swensen J, Shattuck-Eidens D, et al. A strong candidate for the breastand ovarian cancer susceptibility gene BRCA1. 1994;266:120-122.
2. Wooster R, Neuhausen SL, Mangion J, et al. Localization of a breast cancersusceptibility gene, BRCA2, to chromosome 13q12-13. 1994;285:2088-2090.
J Natl Cancer Inst.
3. Jenks S. BRCA1 patent dispute resolved; NIEHS included. 1995;87:412-413.
4. Koenig BA, Greely HT, McConnell LM, Silverberg HL, Raffin TA, and themembers of the Breast Cancer Working Group of the Stanford Program inGenomics, Ethics, and Society. Genetic testing for BRCA1 and BRCA2: recommendationsof the Stanford Program in Genomics, Ethics, and Society. 1998;7:531-545.
5. Myriad Genetics, Inc. September 12, 2002. First ever campaign for cancer predictivetest to market BRACAnalysis in Denver and Atlanta [press release].Available at: http://www.corporate-ir.net/ireye/ir_site.zhtml?ticker=mygn&script=413&layout=9&item_id=333030. Accessed December 27, 2004.
J Clin Oncol.
6. Gray S, Olopade O. Direct-to-consumer marketing of genetic tests for cancer:buyer beware. 2003;21:3191-3193.
7. Gollust SE, Hull SC, Wilfond BS. Limitations of direct-to-consumer advertisingfor clinical genetic testing. 2002;288:1762-1767.
8. Graham J. Gene-test makers woo consumers; pitches exploit fear of cancer, criticscomplain. September 8, 2002.
9. Austin M. Marketing of cancer test exploits fears, critics say. August 30, 2002.
10. Flam F. Gene tests' accuracy reassessed; disease-risk figure may be overstated.September 17, 2002.
West J Med.
11. Pinsky LE, Culver JB, Hull J, et al. Why should primary care physicians knowabout breast cancer genetics? 2001;175:168-173.
J Cancer Educ.
12. Mouchawar J, Klein C, Mullineaux L. Colorado physicians' hereditary breastcancer knowledge and practice behavior. Spring 2001;16(1):33-37.
Cancer Epidemiol Biomarkers
13. Wideroff L, Freedman AN, Olson L, et al. Physician use of genetic testing forcancer susceptibility: results of a national survey. 2003;12:295-303.
14. Escher M, Sappino AP. Primary care physicians' knowledge and attitudestowards genetic testing for breast-ovarian cancer predisposition. 2000;11:1131-1135.
Arch Intern Med.
15. Lerman C, Croyle R. Psychological issues in genetic testing for breast cancersusceptibility. 1994;154:609-616.
Am J Med Genet.
16. Lodder L, Frets PG, Trijsburg RW, et al. Psychological impact of receiving aBRCA1/BRCA2 test result. 2001;98:15-24.
Patient Educ Couns.
17. Meiser B, Butow PN, Barratt AL, et al, and the Psychological ImpactCollaborative Group. Long-term outcomes of genetic counseling in women atincreased risk of developing hereditary breast cancer. 2001;44:215-225.
18. Schrag D, Kuntz KM, Garber JE, Weeks JC. Life expectancy gains from cancerprevention strategies for women with breast cancer and BRCA1 or BRCA2 mutations.2000;283:617-624.
N Engl J Med.
19. Schrag D, Kuntz KM, Garber JE, Weeks JC. Decision analysis—effects of prophylacticmastectomy and oophorectomy on life expectancy among women withBRCA1 or BRCA2 mutations. 1997;336:1465-1471.
J Natl Cancer Inst.
20. Hartmann LC, Sellers TA, Schaid DJ, et al. Efficacy of bilateral prophylacticmastectomy in BRCA1 and BRCA2 gene mutation carriers. 2001;93:1633-1637.
N Engl J Med.
21. Kriege M, Brekelmans CT, Boetes C, et al, and the Magnetic ResonanceImaging Screening Study Group. Efficacy of MRI and mammography for breast-cancerscreening in women with a familial or genetic predisposition. 2004;351:427-437.
22. Myriad Genetic Laboratories, Inc. 2004. Mutation prevalence tables. Availableat: http://www.myriadtests.com/provider/mutprevo.htm. Accessed December 27,2004.
J Clin Oncol.
23. Domchek SM, Eisen A, Calzone K, et al. Application of breast cancer risk predictionmodels in clinical practice. 2003;21:593-601.
24. The BayesMendel lab. BRCAPRO. Available at:http://astor.som.jhmi.edu/brcapro/. Accessed December 27, 2004.
25. Thewes B, Meiser B, Hickie IB. Psychometric properties of the Impact of EventScale amongst women at increased risk for hereditary breast cancer. 2001;10:459-468.
26. Direct to Consumer (DTC) Advertising of Genetic Tests. Workshop summary.The National Human Genome Research Institute. March 23, 2004. Available athttp://www.nhgri.nih.gov/12010660. Accessed August 24, 2004.
27. Direct to Consumer Marketing of Genetic Tests. The National HumanGenome Research Institute. March 23, 2004. Available at: http://www.nhgri.nih.gov/12010659. Accessed August 24, 2004.
28. Myriad Genetic Laboratories, Inc. Integrated Consumer Awareness Campaign:Post-Pilot Evaluation Report. Denver, Colo: July 2004.
29. Absetz P, Aro AR, Sutton SR. Experience with breast cancer, pre-screening perceivedsusceptibility and the psychological impact of screening. 2003;12:305-318.
30. Perez-Stable EJ, Sabogal F, Otero-Sabogal R, Hiatt RA, McPhee SJ.Misconceptions about cancer among Latinos and Anglos. 1992;268:3219-3223.
J Community Health.
31. Laws MB, Mayo SJ. The Latina Breast Cancer Control study, year one; factorspredicting screening mammography utilization by urban Latina women inMassachusetts. 1998;23:251-267.
Am J Prev Med.
32. Chavez LR, Hubbell FA, Mishra SI, Valdez RB. The influence of fatalism on self-reporteduse of Papanicolaou smears. 1997;13:418-424.
Patient Educ Couns.
33. Thompson HS, Valdimarsodottir HB, Jandorf L, Redd W. Perceived disadvantagesand concerns about abuses of genetic testing for cancer risk: differences acrossAfrican American, Latina and Caucasian women. 2003;51:217-227.
34. Centers for Disease Control and Prevention. Genetic testing for breast andovarian cancer susceptibility: evaluating direct-to-consumer marketing—Atlanta,Denver, Raleigh-Durham, and Seattle, 2003. 2004;53:603-606.
FDA Consumer magazine.
35. Lewis C. The impact of direct-to-consumer advertising. March-April 2003. Available at: http://www.fda.gov/fdac/features/2003/203_dtc.html. Accessed August 24, 2004.
Health Aff (Millwood).
36. Weissman JS, Blumenthal D, Silk AJ, et al. Physicians report on patientencounters involving direct-to-consumer advertising. January-June 2004; Suppl Web Exclusives:W4-219-233.
37. Mouchawar J, Alford SH, Laurion S, et al. Impact of direct-to-consumer advertisingfor hereditary breast cancer testing on genetic services at a managed careorganization: a naturally-occurring experiment. 2005;7:191-197.
FDA Consumer magazine.
38. Rados C. Truth in advertising: Rx drug ads come of age. July-August 2004. Available at: http://www.fda.gov/fdac/features/2004/404_ads.html. Accessed August 24, 2004.