American Journal of Medical Genetics 66:l-6 (1996)

Notification of a Family History of Breast Cancer:

Issues of Privacy and Confidentiality
Paula R. Winter, Georgia L. Wiesner, John Finnegan, Dianne Bartels, Bonnie LeRoy,
PingLing Chen, and Thomas A. Sellers
Division of Epidemiology, School of Public Health (P.R.W., J.F., P.-L.C., T.A.S.),Division of Genetics and Metabolism
(B.L.), Institute of Human Genetics (B.L., T.A.S.), and Center for Biomedical Ethics (D.B.), University of Minnesota,
Minneapolis, Minnesota; Department of Genetics and Center for Human Genetics (G.L.W.), Case Western Reserve
University School of Medicine and University Hospitals of Cleveland, Cleveland, Ohio

Little information is available about notify- blood relatives (62%,P < 0.01) were aware of
ing individuals with a family history of can- their family history. 43 respondents (12%)
cer about their risk of cancer. With the re- expressed specific concerns about partici-
cent identification of BRCAl, an important pating in the large genetic follow-up study
predisposition gene for breast and ovarian and 16 comments concerned privacy issues.
cancer, genetic testing is becoming avail- Neither the reasons for participation nor an
able to high-risk women and their families. individual’s concern about developing can-
Some of these individuals may not be aware cer was associated with gender of the re-
of their family history and may be notified of spondent, relationship to the proband, or
their family history by medical personnel or awareness of breast cancer in the family. In-
biomedical investigators. This disclosure terestingly, individuals who were notified
could be detrimental to the individual by about their family history through the large
changing their perception of risk, sense of follow-up study were no more likely than
privacy, or psychosocial well-being. other family members to be more concerned
Members of 544 breast cancer families are about developing cancer.
currently being contacted as part of an epi- Understanding the privacy and psycho-
demiologic follow-up study at the University social issues of family members who are
of Minnesota. Some family members were informed about a family history of breast
unaware of their relative’s diagnosis and cancer may aid in developing appropriate
therefore, notification occurred when they guidelines for notification. Risk notification
were contacted by study personnel. To de- in this setting does not appear to have a sig-
termine the impact of risk notification in nificant impact on these family members.
this context, 376 male and female relatives of 01996 Wiley-Liss, Inc.
160 breast cancer probands were surveyed
to assess their prior knowledge of their fam- KEY WORDS: breast cancer, confidential-
ily history of cancer, issues relating to study ity, family history, biomed-
participation, and their concerns regarding ical ethics, notification
the possibility of developing cancer. Follow-
ing a telephone interview about family his-
tory, family members were administered a
short, open-ended questionnaire. INTRODUCTION
The majority of individuals (82%) were A family history of breast cancer is one of the
blood relatives of the proband and 71% were strongest independent predictors for developing breast
either first- or second-degree relatives. A cancer [Harris et al., 19923. As cancer susceptibility
proportion of blood relatives (24%)were not genes are identified, including B R C A l [Miki et al.,
aware of their family history of breast can- 19941and BRCA2 [Wooster et al., 19941,genetic testing
cer. More blood relatives (76%) than non- for breast cancer [King et al., 1993; Lynch et al., 19931
and various other cancers and conditions [Caplan,
Received for publication December 27, 1994; revision received
19921 can be anticipated. The ethical and psychosocial
October 9, 1995. implications of genetic testing will affect all areas of sci-
Address reprint requests to Dr. Georgia L. Wiesner, Assistant ence, medicine, and the law. Before genetic testing is
Professor, Center For Human Genetics, 11100 Euclid Avenue, fully available, an evaluation of the effect the knowl-
Lakeside 1500, Cleveland, OH 44106-5000. edge will have on individuals and families is required.
0 1996 Wiley-Liss, Inc.
2 Winter et al.
A family history of breast or other cancer is currently To initiate the process of updating and extending the
one factor used to estimate the level of genetic risk for pedigrees from the 1944 baseline study, a t least one in-
developing cancer. With the imminent development of dividual had to be identified from each family. This
molecular diagnostic tests for breast cancer [American initial contact person was located using the last known
Society of Human Genetics Ad Hoc Committee, 1994; address of family members, telephone directories, direc-
Motulsky, 19941, accurate family history and pedigree tory assistance operators, and reverse directories. Once
information will be important. For individuals who are located, the individual was contacted by telephone to
knowledgeable about their family histories, the avail- verify family membership and mailed a letter explain-
ability of testing may offer reassurance or improve ac- ing the study. A staff member then called the initial con-
cess to appropriate medical care [Biesecker et al., tact person for the pedigree extension interview. If the
19931. However, some individuals are unaware that initial contact was unable to provide information for all
they have a family history of cancer and may learn of a family members, another relative was called without an
relative’s diagnosis from individuals outside the family. introductory letter and asked for assistance. The num-
What impact will the disclosure of a family history of ber of individuals contacted per family during pedigree
cancer have on a n individual’s perception of risk or psy- extension was dependent on the family size, geographic
chosocial well-being? Will notification increase a per- location, and whether family members communicated
son’s general anxiety or worry about developing cancer? regularly.
One guiding principle should be lessening potential Notification Questionnaire
harmful effects that disclosure may have on a n individ-
ual. Before genetic testing can be offered on a clinical One hundred sixty consecutive families were sur-
basis, these issues need to be addressed so that we can veyed with the notification questionnaire by telephone
be prepared to offer appropriate genetic counseling and interview from April 30, 1993 to March 3, 1994. Up to
medical management [Lerman et al., 19911. five contacts, or all if less than five, in each family were
We were in a unique position at the University of randomly interviewed at the conclusion of the pedigree
Minnesota to examine the perceptions of individuals af- extension interview without regard to gender or rela-
ter notification of a family history of breast cancer. Be-tionship to the proband. The survey consisted of five
ginning in 1993, a follow-up study of 544 breast cancer questions, four of which were open-ended to facilitate
families was begun a t the University of Minnesota unbiased responses. Questions were designed to assess
[Sellers e t al., 19951. Early in the course of the study,1)each contact’s prior knowledge of his or her family
hereafter referred to as the large follow-up study, some history of breast cancer, 2) reasons for and 3) concerns
participants independently contacted the study person- about participating in the follow-up study, and 4) per-
nel with specific concerns about the notification process sonal concerns about developing cancer. The fifth ques-
[Wiesner et al., 19931. In some cases, individuals were tion permitted each respondent a n opportunity to con-
tribute any additional personal or family information
unaware of a family history of breast cancer. Others
that he or she felt was relevant. For each respondent
questioned the manner in which they were contacted.
the name, relationship to the proband (via family iden-
Therefore, we embarked on a systematic study to ex- tification number) and gender was recorded. Age was
plore the impact of notification of a family history of recorded only for those respondents who were eligible
breast cancer on these family members. We wished to for follow-up in the large follow-up study (i.e., first- or
define the proportion of family members that were second-degree female relatives or spouses of first- or
knowledgeable about their family history, assess rea- second-degree male relatives).
sons for participation, and evaluate issues of privacy In order to assess whether the survey was represen-
and confidentiality in this cohort. tative of this cohort’s overall concerns for privacy and
METHODS confidentiality, individuals from 53 consecutive fami-
Study Families lies were monitored for spontaneous comments during
the pedigree extension interview andlor administration
In 1944, a case-control family study was initiated a t of the notification questionnaire. Concerns about par-
the University of Minnesota to investigate the influ- ticipation expressed during the notification survey
ences of childbearing, breastfeeding, and positive fam- were considered separately from all other spontaneous
ily history on the occurrence of breast cancer [Anderson comments. Staff members noted all spontaneous or un-
et al., 19581; 544 index cases with breast cancer from solicited comments about privacy or confidentiality is-
the state of Minnesota were ascertained between 1938 sues, the nature of the comments, and the gender and
and 1952 at the Tumor Clinic without prior knowledge degree of relationship of the participant to the proband.
of their family history. A large follow-up study of these
families is currently being conducted [Sellers et al., Data Analysis
19951. Eligible participants include first- and second- Responses to each question were reviewed and as-
degree female relatives of probands (blood relatives) signed to logical categories. One respondent was ex-
and female spouses of probands’ first- and second- cluded from analysis due to incomplete personal infor-
degree male relatives (marry-ins). Data on disease sta- mation. Statistical evaluation of the data was performed
tus and cancer risk factors are collected by a combina- by means of contingency table analyses. In addition, the
tion of telephone interviews, mailed questionnaires, specific cases who independently contacted study per-
and medical records reviews. sonnel were excluded from the analysis.
 Impact of Notification of a Family History of Breast Cancer 3
TABLE I. Gender and Biological Relationship of 376 Family than their male counterparts (64% of grandsons and
Members Surveyed With the Notification Questionnaire 62% of nephews).
Frequency (n) Percent (%) Motivation for participation in the larger follow-up
study was given by 375 respondents (429 total re-
Gender sponses). Eleven categories were defined (Table 111),
Male 78 21 with the vast majority of responses (73%) classified a s
Female 298 79
Degree of relationship to proband “general or altruistic” reasons. While a higher propor-
Blood relative (total) 308 82 tion of women than men stated a specific reason for par-
Proband 2 1 ticipating (13% vs. 5%, P = 0.041, all other responses
Sibling 7 2 were similar for men and women. A history of cancer in
Child 73 19 relatives, close friends, or the participants themselves
Grandchild 45 12 was cited by 54 (15%) respondents as a reason for par-
Niecehephew 140 37
Third degreeldistant 41 11 ticipating, but there was no significant difference be-
Non-blood relative 68 18 tween individuals who were previously aware of their
family history and those who were unaware. A similar
proportion of blood (12%) and non-blood relatives (10%)
stated that family history was a reason to enter the
RESULTS large follow-up study. Among blood relatives, 29% of
Most of the 376 respondents were female and biolog- siblings (n = 2), 16% of children (n = 12), and 12% of
ically related to the proband (Table I). Among blood rel- distant relatives (n = 5) cited family history of cancer
atives, 21% were first-degree relatives (children and compared to only 9% of nieces and nephews (n = 13)
sisters) and 49% were second-degree relatives (grand- and 9% of grandchildren (n = 4).
children, nieces, nephews). Non-blood relatives in- A total of 366 individuals responded to a n open-
cluded all individuals who had married into the family ended question regarding additional participation con-
(marry-ins), caregivers for family members, adopted cerns; the majority (88%)had no comments (Table 11).
relatives, or “step-” relatives. The mean age of 258 There was little difference in the frequency of concerns
blood relatives for whom age was recorded was 66.0 by gender or biological relationship to the proband. Of
years, a s compared with the mean age of 64.4 years for 301 responding blood relatives, 33 (11%) had participa-
65 non-blood relatives. A fraction of those interviewed tion concerns compared to 15% of 65 responding non-
(26%) were unaware of their family history, with little blood relatives. The 43 specific participation issues fell
difference observed based on the gender of the respon- into two broad categories: worries about the amount of
dent (Table 11). As expected, blood relatives were sig- time and level of effort required (n = 27) and concerns
nificantly more likely than non-blood relatives to be regarding privacy or confidentiality (n = 16). The ma-
aware of a family history of breast cancer on the jority of individuals with either concerns about effort
proband’s side of the family (77% vs. 62%, P < 0.011, and time (n = 19, 70%) or privacy and confidentiality
and those who were less closely related to the proband (n = 14, 88%) were aware of a family history of breast
were less likely to be aware of a family history of breast cancer. In addition, six respondents who agreed to com-
cancer. Among children of probands, the proportion of plete the notification survey refused to participate in
sons and daughters that reported awareness were sim- the large follow-up study. The reasons for refusal were
ilar (100% and 94%, respectively). However, female sec- primarily about time and effort rather than privacy.
ond-degree relatives (77% of granddaughters and 77% The level of concern about developing cancer was high
of nieces) were more aware of the breast cancer history in all groups (Table 11).Interestingly, the concern was as

TABLE 11. Knowledge of Family History, Participation Concerns, and Concerns

About Cancer by Gender and Relationship to the Proband in Family Members of
160 Breast Cancer Probands
Aware of family Participation Concern about
history concern developing cancer
(n = 375) (n = 366) (n = 372)
(n) (%I (n) (%I (n) (%I
Gender
Male 58 75 9 12 50 66
Female 218 73 34 12 193 65
Degree of relationship to proband
Blood relative 234 76 33 11 203 67
Proband 2 100 1 50 1 50
Sibling 7 100 1 17 5 71
Child 70 96 8 11 57 78
Grandchild 33 73 3 7 33 73
Niecehephew 103 74 17 13 76 55
Third degreeldistant 19 46 3 7 31 78
Non-blood relative 42 62 10 15 40 60
4 Winter et al.
TABLE 111. Distribution of Motives Given by 375 Family Members for Their
Participation in a Genetic Epidemiologic Study of Breast Cancer
Frequency (n) Percent
Motivation
~~~ ~
(n = 429)” (%Ih
GeneraValtruistic 274 73.1
Family history of cancer 43 11.5
Seriousness of cancer/ media coverage/ health professional 21 5.6
Another relative is participating 16 0.4
To gain knowledge about breast cancer 15 4.0
Good reputation of University of Minnesota 13 3.5
Personal history of cancer 7 1.9
Close friend had cancer 4 1.1
Family history of another disease 4 1.1
No reason specified 26 6.9
Refused participation in genetic epidemiological study 6 1.6
a Number of responses is greater than participants because of multiple responses from some participants.
Percents may not sum to 100%because of multiple responses from some participants.

high in male as in female respondents, and non-blood a n impact on a n individual’s decision to enter the study.
relatives were only slightly less concerned than blood Respondents who expressed such concern were some-
relatives. When asked “Is the possibility of getting can- what less likely to have specific issues about participat-
cer a health concern for you and your family?”, 358 in- ing in the larger follow-up study than individuals with-
dividuals answered with one comment and 15 individu- out cancer concerns (10% vs. 15%,P < 0.15).
als offered two (Table IV). The most frequent comment In order to determine the level of concern for issues
involved a general level of concern with statements such relating to privacy or confidentiality, the interviews for
as “everyone should be concerned” and “I’m aware of it, 53 families were monitored for unsolicited comments.
but I don’t dwell on it.” Prior knowledge of their family There were a total of 630 contacts in 53 consecutive in-
history of cancer (15%) or personal history of cancer terviews to extend family pedigrees. Only 19 individu-
(6%) were cited infrequently. Additional analyses were als (3%) in 15 families expressed some concern about
performed to assess whether the degree of biological re- privacy. Seven comments occurred during the adminis-
latedness or prior knowledge of a family history of can- tration of the notification questionnaire, and 12 (2%)
cer was associated with a n increase in anxiety about de- were spontaneously offered in the pedigree extension
veloping cancer. Interestingly, the level of concern interview. Frequency of privacy concerns did not vary
expressed by the proportion of third-degree or more dis- by gender or relationship to the proband. The specific
tant relatives was similar to if not greater than the pro- content of six spontaneous concerns (50%)compared to
portion of first- and second-degree relatives (Table 11). a total of eleven questionnaire-assessed concerns (69%)
While the number of cancer concerns expressed by sons involved general unwillingness to provide personal in-
(77%) and daughters (79%) was similar, 82% of grand- formation over the phone. Comments addressing fam-
daughters (n = 28) and only 46% of grandsons (n = 5) ily dynamics or sensitivity for the privacy of a particu-
reported being concerned. Furthermore, prior knowl- lar relative were spontaneously offered by one son, two
edge of family history of cancer was unrelated to grandchildren, two third-degree relatives, and one non-
concern (76% vs. 69%, P = 0.14). Similar comparison blood family member. Two male relatives sponta-
of family history awareness and cancer concern by gen- neously asked whether their relatives’ or their own
der revealed no significant differences. We wondered names and addresses would be made commercially
whether concern about developing cancer would have available for companies to send unsolicited advertise-

TABLE IV. Distribution of Responses by 373 Family Members

t o Open-Ended Questions About Cancer Concerns
Frequency (n)” Percent
Comment (n = 388) (%Ih
General concern 121 32.4
Health issues (regular checkups, healthy lifestyle) 71 19.0
Family history of cancer 55 14.7
Personal history of cancer 21 5.6
Family history of another disease 13 3.5
Widespread media coverage of cancer 7 1.9
Privacy issue (insurance) 1 0.3
No concern specified 99 26.5
Three participants declined to respond t o this question and 15 offered 2 responses.
% may not sum to 100%because of multiple responses from some participants.
 Impact of Notification of a Family History of Breast Cancer 5
ments. During administration of the notification ques- most of the privacy concerns were expressed by partic-
tionnaire, two male respondents expressed concern ipants already aware of their family history, and it was
about the release of data t o insurance agencies. encouraging to find a small number of privacy concerns
among individuals newly notified of their family his-
DISCUSSION tory of breast cancer. The majority of individuals who
This report describes the impact of notification of a expressed concerns were skeptical of sharing personal
family history of breast cancer on 376 members of 160 information over the telephone with unknown callers.
families originally ascertained by a single proband with Other worries involved fear of adverse personal conse-
breast cancer. Nearly one quarter of our cohort was un- quences associated with insurance companies or re-
aware of a family history of cancer prior to contact by ceiving “junk mail.” Very few privacy concerns dealt
our study. We had hypothesized that notification of a with personal feelings, overt suspicion, or the privacy
family history might increase an individual’s percep- or feelings of other family members. Thus, if privacy
tion of his or her risk for developing cancer and lead to concerns were raised by notification in our epidemio-
a higher level of anxiety. If this were true, we would logic research context, such concerns were not neces-
have expected to see a greater level of concern in blood sarily overt or immediate.
relatives who had previously been unaware of their The ethical issues encountered with the disclosure of
family history; such an outcome was not observed. a family history of breast cancer are complex because
An unexplainably high proportion of third-degree or notification could have a negative impact on the indi-
more distant relatives expressed concern about devel- vidual at a higher risk for developing cancer [Lerman
oping cancer. One may question whether the cancer et al., 19911. However, the initial effect of notification
concern indicated by participants in this study stems on the 376 family members of 160 breast cancer fami-
from their family history, as spontaneously suggested lies was minimal in this study. We found no evidence for
by 55 respondents, or whether the increasing role of the a detrimental effect of notification when we analyzed
media in promoting cancer screening is the primary in- the responses from the individuals who were unaware
fluence, as mentioned by 7 respondents. Furthermore, of their family history. While results of our study are
a respondent’s perception of a family history of cancer encouraging regarding the initial responses following
may be substantially different from this study’s defini- notification, the results should be interpreted cau-
tion of family history (a diagnosis of breast cancer in tiously. First, the response t o notification of a family
the proband). A small number of respondents clearly history of cancer in a distant relative could be quite dif-
indicated awareness of the proband’s diagnosis but ex- ferent from the notification that a close relative is car-
pressed the opinion that they did not have a history of rying a known cancer-predisposing gene. Second, the
breast cancer in their family. sample of families studied were not selected on the ba-
There was essentially no difference between the pro- sis of a strong family history of breast cancer, merely
portion of males and females who expressed concern through a single proband with the disease. Thus, only a
about developing cancer. However, it is unclear minority of the families will ultimately be shown to be
whether the men who expressed a cancer concern were “hereditary.” Future studies should address the poten-
more worried about their own health or about the
tial long term effects of notification by monitoring be-
health of their female relatives. Little research exists
havioral responses to cancer screening after disclosure.
on cancer risk perception among male relatives of
breast cancer patients. Many studies have demon- In addition, the psychosocial effects of notification
strated that female relatives, especially daughters, of should be included in future studies of the impact of no-
breast cancer patients tend to greatly overestimate tification on individuals at genetic risk for cancer.
their risk of developing breast cancer [Biesecker et al.,
ACKNOWLEDGMENTS
1993; Lynch et al., 1993; Wellisch et al., 19911, while
other women a t high risk for breast cancer greatly un- We are greatly indebted to the family members who
derestimate their risk [Lynch et al., 1993; Thirlaway made this study possible. This work was supported in
and Fallowfield, 1993; Kash et al., 19921. Evans et al. part by NIH grant R01 CA55747.
[19931found that a woman’s personal estimation of de-
veloping breast cancer increases with significance of REFERENCES
family history. Our results indirectly support this find- American Society of Human Genetics Ad Hoc Committee (1994):
ing because the nieces and non-blood relatives in our ASHG policy statement on genetic testing for breast and ovarian
sample reported a lower frequency of cancer concern cancer predisposition. Am J Hum Genet 55:i-iv.
than first-degree relatives and granddaughters. How- Anderson V, Goodman H, Reed S (1958):“Variables Related to Human
Breast Cancer.” Minneapolis, Minnesota: The University of Min-
ever, we did not ask the participants to estimate their nesota Press.
risk or specify a level of concern. Biesecker B, Boehnke M, Calzenek K, Markel D, Garber J, Collins F,
We were interested in the potential impact that noti- Weber B (1993): Genetic counseling for families with inherited sus-
fication might have on participation in the study and on ceptibility to breast and ovarian cancer. JAMA 269:1970-1974.
the respondents’ perceptions of privacy. Participants Caplan A (1992): Mapping Mortality: Ethics and the Human Genome
Project: “If I Were a Rich Man Could I Buy a Pancreas?” Indiana
who had no previous knowledge about a family history University Press, pp 118-142.
of breast cancer might have been expected t o have more Evans D, Burnell L, Hepwood P, Howell A (1993): Perception of risk in
privacy issues or concerns about entering the study; women with a family history of breast cancer. Br J Cancer 67:
our data do not support this premise. Unexpectedly, 612-614.
6 Winter et al.
Harris J, Lippman M, Veronesi U, Willet W (1992): Breast cancer Motulsky, A (1994): Predictive genetic diagnosis. Am J Hum Genet
(first of three parts). N Engl J Med 327:319-328. 55603-605.
Kash K, Holland J , Halper M, Miller D (1992): Psychological distress Sellers T, Anderson V, Potter J, Bartow S, Chen P, Everson L, King R,
and surveillance behaviors of women with a family history of Kuni C, Kushi L, McGovern P, Rich S, Whitbeck J, Wiesner G
breast cancer. J Natl Cancer Instit 84:24-30. (1995): Epidemiologic and genetic study of 544 Minnesota Breast
Cancer Families: Design and methods. Genet Epidemiol 12:
King M, Rowel1 S, Love S (1993): Inherited breast and ovarian cancer: 417- 429.
What are the risks? What are the choices? JAMA 269:1975-1980.
Thirlaway K, Fallowfield L (1993):The psychological consequences of
Lerman C, Rimer B, Engstrom P (1991): Cancer risk notification: Psy- being a t risk of developing breast cancer. Eur J Cancer Prevention
chosocial and ethical implications. J Clin Oncol9:1275-1282. 2:467- 471.
Lynch H, Watson P, Conway T, Lynch J, Slominski-Caster S, Narod S, Wellisch D, Gritz E, Schain W, Wang H, Siau J (1991): Psychological
Feunteun J, Lenoir G (1993): DNA screening for breadovarian functioning of daughters of breast cancer patients: Part I: Daugh-
cancer susceptibility based on linked markers. Arch Int Med 153: ters and comparison subjects. Psychosomatics 32:324-336.
1979-1987. Wiesner GL, Winter R, Bartels D, Sellers TA, LeRoy BS (1993):Com-
Miki Y, Swensen J, Shattuck-Eidens D, Futreal PA, Harshman K, plex issues encountered with the disclosure of a family history of
Tavtigian S, Liu Q, Cochran C, Bennett LM, Ding W, Bell R, breast cancer. Am J Hum Genet 53:1511 (abstract).
Rosenthal J , Hussey C, Tran T, McClure M, Frye C, Hattier T, Wooster R, Neuhausen SL, Mangion J , Quirk Y, Ford D, Collins N,
Phelps R, Haugen-Strano A, Katcher H, Yakumo K, Gholami 2, Nguyen K, Seal S, Tran T. Averill D, Fields P, Marshall G, Narod
Shaffer D, Stone S, Bayer S, Wray C, Bogden R, Dayananth P, S, Lenoir GM, Lynch H, Feunteun J, Devilee P, Cornelisse CJ,
Ward J, Tonin P, Narod S, Bristow PK, Norris FH, Helvering L, Menko FH, Daly PA, Ormiston W, McManus R, Pye C, Lewis CM,
Morrison P, Rosteck P, Lai M, Barrett JC, Lewis C, Neuhausen S, Cannon-Albright L, Peto J , Ponder BAJ, Skoinick MH, Easton DF,
Cannon-Albright L, Goldgar D, Wiseman R, Kamb A, Skolnick MH Goldgar DE, Stratton MR (1994): Localization of a breast cancer
(1994): A strong candidate for the breast and ovarian cancer sus- susceptibility gene, BRCA2, to chromosome 13q12-13. Science
ceptibility gene BRCA1. Science 266:66-71. 265~2088-2090.