Familial Cancer 2: 159–168, 2003. 2003 Kluwer Academic Publishers. Printed in the Netherlands.
What motivates interest in attending a familial cancer genetics clinic? L. Fraser1, S. Bramald2, C. Chapman2, C. Chu3, V. Cornelius4, F. Douglas5, A. Lucassen6, A. Nehammer7, S. Sutton8, M. Trivella4 and S. Hodgson9 1
Department of Medical & Molecular Genetics, Guy’s Hospital, London, UK; 2 Department of Clinical Genetics, The Churchill Hospital, Oxford, UK; 3 Department of Clinical Genetics, St James’s University Hospital, Leeds, UK; 4 Cancer Research UK, Medical Statistics Group, Oxford, UK; 5 Department of Clinical Genetics, Royal Victoria Infirmary, Newcastle, UK; 6 Department of Clinical Genetics, Princess Anne Hospital, Southampton, UK; 7 Department of Clinical Genetics, Guy’s Hospital, London, UK; 8 Department of Public Health and Primary Care, University of Cambridge, UK; 9 Department of Clinical Genetics, St George’s Hospital, London, UK Received 6 August 2003; accepted in revised form 2 September 2003
Key words: cancer genetics services, familial cancer, familial cancer clinics, genetic counselling, reasons for attending
Abstract The motivation of people who seek advice about a family history of cancer was explored in a cross sectional study of new cancer referrals to five regional cancer genetics centres in England: the PACT (patient and clinical team) psychosocial study. One hundred sixty-two people took part. Measures were source of referral, estimated and perceived cancer risk, level of cancer worry, and personal and family-centred reasons for wanting to be seen in clinic. General practitioners referred more people than hospital doctors, and referred a larger proportion of people at low genetic risk of developing cancer. More than half of the participants had been the first to raise the issue of their family history of cancer. Personal motivation for referral is clearly different for those who have had a diagnosis of cancer and for those with children, compared to unaffected and childless people, and is characterised by altruistic concern for other family members rather than a perception of increased personal risk. Men and people from ethnic minorities are very significantly under-represented. Understanding people’s motivation may be useful in targeting genetic counselling for people with a family history of cancer. Introduction In 1998, a working group for the Chief Medical Officer published a report on the provision of cancer genetics services [1]. One of the report’s main recommendations was the triage of people seeking advice about their family history of cancer to different levels of health care depending on their risk of developing cancer. According to this policy, when the pattern of cancers observed in a family history suggests the presence of a single gene disorder conferring high cancer risk, the person should be referred to a specialist familial cancer genetics centre. If the level of risk is lower but merits screening for cancer (moderate risk), they could be seen at a hospital cancer unit. For those at population or only slightly increased risk (lower risk), reassurance can be provided by general practitioners.
A national survey carried out in the same year described the state of service provision at cancer genetics centres across the United Kingdom [2]. In addition to reporting regional inequalities in access to specialist care (too few cancer genetics staff, and unavailability of some cancer screening and genetic testing services), it was apparent that over a quarter of referrals were for people in the lower risk group. Familial cancer clinics in the United Kingdom are experiencing a surge in referrals. Cancer Research UK reports that family cancer clinics with support from them have been ‘inundated’ with women wanting to find out if their family history of breast cancer increases their personal risk of the disease [3]. Each clinic received about 500 referrals in 2000. Since this is an important use of healthcare resources, information is needed about why referrals are made and by whom, and whether
Correspondence to: L. Fraser, Gynaecology and Oncology Unit, Barts and The London, Queen Mary’s School of Medicine and Dentistry, Charterhouse Square, London EC1M 6GR, UK. Tel: +44-20-78825952; Fax: +44-20-78825958; E-mail:
[email protected]
160 specific socioeconomic or ethnic groups dominate referrals, in order to plan for future service development. Recent indications are that some ethnic minority groups and those with limited education are under-represented among family history clinic attendees [4]. The motivation of people interested in attending a family cancer clinic is not well understood and requires clarification. Activity is currently dominated by referrals from breast cancer families [2]. A survey of unaffected women with a family history of breast cancer waiting to be seen at a familial breast cancer clinic appears to reflect a perception of increased personal risk [5]. Brain et al. reported that almost a third of their moderate to high risk sample said the main reason for wanting to be seen in clinic was to obtain information about their own risk of developing breast cancer. At the heart of genetic counselling for inherited cancer risk is a willingness to talk about cancer in the family. It is anticipated that referral of families with a history of colorectal cancer will increase as knowledge about the genetic component of some colorectal cancers becomes more widely recognised. Collins et al. have looked at the motivation of men and women waiting for an appointment at a familial colorectal cancer clinic [6]. They reported that 75% of their sample believed that establishing whether there is a gene for cancer susceptibility in the family is more important than any other reason for coming to the clinic. Other reasons known to motivate interest in attending a family cancer clinic include concern about the risk to other family members (notably children and siblings), to obtain information about cancer prevention, access to screening, and to reduce anxiety [7–10]. The motivation of people who initiate a referral may be different from those referred at the suggestion of someone else. In many cases, people raise the issue of the family history of cancer as they approach the age at which a close relative was diagnosed with cancer, or the age at which they died [11]. The PACT (patient and clinical team) psychosocial study describes the pattern of new cancer referrals at five regional cancer genetics centres in England. It assesses the source of referrals and explores why people want to be seen at a familial cancer clinic. There is limited information in the literature about the motivation of people interested in attending a family cancer clinic. Whilst there may be a wide variety of reasons for interest in attending, this study aims to describe people’s motivation according to different grouping variables reflecting the broad range of people being seen. These include who instigates referrals, cancer status, having children, perceived risk of developing cancer, level of cancer worry, and predominant type of cancer in the family history. The knowledge gained will be useful in tailoring genetic counselling to meet people’s needs [12].
L. Fraser et al. Participants and procedure The data were gathered at five genetics centres receiving support from Cancer Research UK in England over the same eight-week period at four centres and over four weeks at a fifth (Centre 1) in 2001. Each centre runs a familial cancer genetics clinic. Appointments were offered to all referrals at one centre (Centre 1) irrespective of the level of cancer risk, whilst the others were more restrictive, aiming to prioritise high risk referrals. People were recruited before any triage of referrals. We used an opt-in method of recruitment in an attempt to avoid provoking anxiety in people waiting to be offered their first clinic appointment, which may have reduced attendance rates. Opt-in methods are known to depress participation rates [13]. A consultant at each centre wrote to notify newly referred people about the study, enclosing a reply slip for those interested in taking part. No further contact was made with those declining or not responding. A study pack was posted to respondents. They were asked to complete and return the study questionnaire before their first clinic appointment. Pre-paid envelopes were supplied at both stages of contact. Four centres sent reminder letters to initial responders. Information about family history of cancer and estimated genetic risk of developing cancer was assessed by clinicians at each centre. Sources of information used to estimate risk included the referral letter, the clinic’s own family history questionnaire, a telephone interview, home visit, clinic consultation, or other source to confirm diagnoses (cancer registry, histopathology report, death certificate). Non-linked, anonymised group data (age, gender, and type of cancer predominant in family history) were gathered for people referred during the study period who did not take part. Ethical approval for the study was obtained from the South East Multi-centre Research Ethics Committee.
Measures Sociodemographic characteristics Information about personal history of cancer (i.e. unaffected or affected) was collected, along with information on age, gender, ethnicity, educational level, marital status, number of children and urban/rural residence. Three indicators of social deprivation were assessed: employment status, home ownership, and car/van ownership [14]. Enumeration district data from the 1991 census were allocated to patients on the basis of their postcode and averaged across centres to give proxy sociodemographic variables for each centre, using the Townsend score [15, 16]. Negative values reflect less deprived areas and positive values reflect more deprived areas. The average is 0.
What motivates interest in attending a familial cancer genetics clinic? Family history Details about the number of family members diagnosed with cancer were recorded. Participants were categorised to cancer type according to the predominant type of cancer occurring in the family history: breast, ovarian, colorectal, and other. Referral patterns The sources of referral, who had first raised the issue of a family history of cancer and who had instigated the referral were noted. Estimated risk Guidelines for estimating the genetic risk of developing cancer were provided to clinicians: ‘lower risk’, population level/marginally increased; ‘moderate risk’, sufficient to merit screening/surveillance; ‘high risk’, likely to be a single gene disorder in the family [2]. Perceived risk of cancer Participants were asked to identify a member of their family diagnosed with cancer who they felt had the greatest impact on their own perceived risk of developing cancer: the ‘index relative’. Subjective risk perception was assessed by asking respondents to indicate what they felt their chance was of getting cancer in their lifetime (of if they have been diagnosed with cancer, of getting it again) compared to the average person in the population. Response options were: lower than average, about average, moderately increased above average, much higher than average [17].
161
asked to rate their level of agreement on the personal relevance of each statement on a 5-point rating scale (1 = strongly disagree to 5 = strongly agree). Participants were also asked to nominate one reason from the list as their ‘main’ reason for wanting to attend. Statistical methods Most statistical analyses were carried out using SPSS v 11.5. Standard descriptive techniques were used to examine the demographic and medical characteristics of the participants, and the sources of referrals. Chi-square tests were used to assess group differences for categorical variables and t-tests for continuous variables. The Mantel–Haenszel chi-square test was used to explore the linear association between the source of referral and level of estimated risk. Mean weighted scores were calculated to describe the level of agreement for reasons given by participants for their interest in attending a family cancer clinic. Independent t-tests and one-way analysis of variance (ANOVA) were used to compare reasons across groups, and significant main ANOVA effects were explored with post hoc Tamhane tests. The Pearson’s correlation coefficient was calculated to explore indications of association between participant’s age and the age when the ‘index relative’ developed cancer. The degree of agreement between participants’ perceived level of risk and clinicians’ estimated risk were assessed with the weighted Cohen’s kappa test of agreement [20], using STATA v 7.0. Kappa scores below 0.0 indicate ‘poor’ agreement, 0.0 to 0.20 = ‘slight’, 0.21–0.40 = ‘fair’, 0.41–0.60 = ‘moderate’, 0.61–0.80 = ‘substantial’, 0.81–1.0 = ‘almost perfect’.
Results Cancer-related worry The extent to which people worry about developing cancer was assessed with a modified version of the breast cancer worry scale, a 6-item scale measuring frequency of concerns about developing cancer and the impact of cancer worry on mood and daily functioning [18, 19]. The wording of the scale was changed to apply to all types of cancer. Scores range from 6 to 24, with higher scores indicating a higher level of cancer-related worry. The internal consistency of the scale was found to be satisfactory (α = 0.87). Motivation to attend family cancer clinic Reasons for wanting to attend a familial cancer clinic were assessed with eleven categorical statements including knowledge about the family history of cancer, a desire to establish personal risk and the risk to other family members, to help the family, reassurance, for advice about risk factors, how to reduce personal risk, prevention and genetic testing, access to screening and following the advice of their GP [5]. Participants were
A total of 477 new cancer referrals were received over the five centres in the study period. Two hundred and seventy nine people (59%) expressed interest in taking part in the study, 154 (32%) did not respond and 44 (9%) declined to take part (a response rate of 68%). Self-reported reasons for declining were recorded (n = 20) and included not having enough time (or being too unwell) to complete the questionnaire, choosing not to think about the health implications of cancer in the family, being unable to cope because of family problems or a recent cancer-related illness or death in the family. One hundred and seventy seven completed study questionnaires were returned. Fifteen were excluded because the questionnaire had been completed after the clinic appointment, or for failing to return a signed consent form (despite reminders). One hundred sixty-two people took part (a participation rate of thirtyfour percent). None of the participants came from the same family.
162
L. Fraser et al.
Sociodemographic characteristics
Family history
The sociodemographic characteristics of the study population are shown in Table 1. Participants were aged between 21 and 70 years (mean 44.3, SD 11.9). The great majority were white, female, had not been previously diagnosed with cancer, were married or cohabiting, with children, educated up to the age of 16 years or beyond, were employed full or part-time, lived in their own home and owned a car or a van. Information about social deprivation was extracted by ward level from the 1991 Census data. One hundred and nineteen respondent postcodes out of the study sample of 162 (73.45%) were matched with the census enumeration district records. The mean social deprivation scores for study participants at three of the five centres were above average (Centre 2 mean –0.3, SD 3.4; Centre 4 mean –0.7, SD 2.2; Centre 5 mean –1.1, SD 2.1) and at two they were below average (Centre 1 mean 1.3, SD 4.6; Centre 3 mean 1.1, SD 3.5).
Participants were categorised according to the predominant type of cancer in their family history: 61% breast cancer, 21% bowel cancer, 14% ovarian cancer, 4% other specified types of cancer (1 prostate cancer and 5 mixed cases of cancer). Participants had an average of 1.6 first degree relatives with cancer (range 0 to 9, mode = 1) and 2.2 second degree relatives (range 0 to 7, mode = 2). The relationship between the participant’s age and the age at which the index relative was diagnosed with cancer was analysed separately for unaffected and affected (previous diagnosis of cancer) participants. There was a small significant positive correlation for unaffected participants (r = 0.262, P ≤ 0.005), but not for those already affected (r = 0.235, P = 0.220).
Table 1. Characteristics of the study sample. N Ethnicity White Non-white
162
Gender Male Female
162
Marital status Married/cohabiting Single Divorced/separated Widowed
162
Education Left school before age 16 years Secondary education (16 to 18 years) Further education (19+ years)
155
Cancer status Affected Unaffected
162
One or more children Yes No
162
Employment Full time work Part time work Not in paid work
161
Car/van ownership Yes No
159
Household Own home Renting Shared ownership/other
162
%
97.5 02.5 07.4 92.6 80.0 08.0 09.0 03.0 15.5 56.8 27.7 23.0 77.0 78.0 22.0 34.2 36.6 29.2 90.6 09.4 77.7 17.3 05.0
Anonymised group data The anonymised group data on age, gender and predominant type of cancer in the family history were added to the study data, giving overall figures for the total number referred during the study period. The study sample was broadly representative of the total referred on these parameters. There was no difference in age (the mean for both groups was 42 years), or the proportion of women (93% participants vs 90% total). The type of cancer predominant in the family history was also very similar: breast–ovarian cancer (75% participants vs 69% total), colorectal cancer (21% vs 20%), and other types of cancer (4% vs 11%). Referral patterns More participants were referred by GPs (52%) than by hospital doctors (38%). The remainder came from another genetics centre (2.5%), were self-referred (2.5%) or referred by a family member (2%), nurse (2%) or other person (1%). More than half reported that they had been the first to raise the issue of the family history of cancer themselves (55%), followed by a family member (20%), hospital doctor (14%), GP (9%), and 2% from a friend, university professor, hospital/breast care nurse. The majority said the referral had been instigated by their GP (39%) or a hospital doctor (33%). Sixteen percent said the referral had been at their own instigation and 6% following a suggestion from a family member. Other reasons given (6%) were: friend, breast care clinic and breast care/clinic/research nurse. Overall, 57.5% of referrals were from urban areas and 42.5% from rural areas. GP referrals for participants living in urban areas formed the largest proportion of referrals (33%). Referrals from urban areas were more likely to be for people at moderate to high risk (χ2 = 5.790, P = 0.055) compared to those from rural areas who were more likely to be at lower risk. Participants living in urban areas were significantly more socially
What motivates interest in attending a familial cancer genetics clinic?
163
deprived than those living in rural areas (t = 4.355, P ≤ 0.001). Estimated risk A quarter of participants were estimated to be at lower risk, 44% at moderate risk, and 31% at high risk. Most estimates were made after direct contact with the participant (clinic consultation 57%, telephone consultation 4%, and home visit 1%). Risk estimates for participants not seen in clinic at the time data collection ceased were based on available information: 21% were made from the family history questionnaire and referral letter, and 17% from the referral letter alone. The proportion of participants at different levels of estimated risk varied across centres (Figure 1) and cancer type (Figure 2). Most participants were from families with a predominant history of breast cancer at high (21%), moderate (25%) and lower genetic risk (13%). The proportion of referrals made by GPs and hospital doctors for a predominant family history of breast cancer (61% vs 57%) and ovarian cancer (18% vs 13%) were similar (Figure 3). However, GPs referred a smaller proportion for colorectal cancer (16% vs 28%). GPs referred significantly more people at lower risk (33%) compared to hospital doctors (17%) (χ2 = 6.776, P ≤ 0.05). This was particularly the case for breast cancer (17% GP vs 6.5% hospital doctor) and ovarian cancer (12% vs 8%). The lower the level of estimated risk, the more likely the referral was to have been made by a GP (M-H χ2 = 6.723, P ≤ 0.01).
Figure 1. Percentage of lower, moderate and high risk estimates by centre.
Figure 2. Percentage of lower, moderate and high risk estimates by cancer type.
Accuracy of risk perception Most participants perceived their risk as moderately increased above average (68%), 31% as much higher than average and only 1% as lower than/about average. Participants’ subjective perceptions of risk were compared with clinicians’ objective risk estimates. Complete data was available for 158 cases. About half were accurate in their perception of risk (48%), with a similar proportion either under-estimating (24%) or over-estimating (28%). This is regarded as a ‘fair’ level of agreement (κ = 0.2154). Cancer worry One hundred sixty respondents provided data on cancer worry. They were initially allocated to one of four categories: no worry (score of 6), low (7 to 12), moderate (13 to 18) and high (19 to 24). The majority of respondents reported low (51%) or moderate (41%) levels of cancer worry. Few reported no worry (2%) or high levels of worry (6%). The data were dichotomised into two categories in order to facilitate analysis: no/low (n = 85) and moderate/high (n = 75).
Figure 3. Proportion of referrals made by GPs and Hospital Doctors by cancer type and estimated level of risk.
Motivation to attend family cancer clinic Main reasons for interest in attending clinic given by unaffected participants (n = 125) are as follows: to establish personal risk (35%), awareness of family history (18%), access to screening (15%), establish risk to family members (10%), reduce personal risk (9%), to help family (4%), information about genetic testing (4%), cancer prevention (2%), reassurance (2%), advised (1%). Mean ratings of agreement for each reason were calculated across respondents (n = 159–162). The number of scores contributing to each mean varied and the range of scores (minimum, maximum) are described for each grouping variable. The reasons are arranged in rank order in Figure 4. Overall, knowledge about the family history of cancer and a desire to establish personal risk of developing cancer achieve the highest ratings of agreement, followed by interest in genetic testing,
164
L. Fraser et al.
Figure 4. Mean ratings of agreement with reasons for wanting to attend (* P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001).
access to screening, and advice about risk factors. Data on four of the grouping variables are plotted in Figures 4a–d to aid interpretation. Significant group differences in mean ratings of agreement for each reason are shown in Tables 2 and 3. Cancer affected status The highest rating of agreement for affected respondents (n = 33–36) was to establish the risk to other family members and to help the family (Figure 4a). Unaffected respondents were more concerned about reducing their own personal risk and were also more likely to have been responding to advice from their GP. Having children The obvious difference between those with (n = 123–127) and without (n = 34–35) children is the per-
ceived importance of establishing the risk of developing cancer for other family members and helping the family (Figure 4b). Respondents with children are more likely to agree that these two reasons are important in their motivation for wanting to be seen at a familial cancer clinic, and for gaining access to screening and advice about cancer prevention. Cancer worry The motivation of people with moderate to high levels of cancer worry (n = 73–75) differs markedly from those who worry less (n = 84–85) (Figure 4c). Those with higher scores on the cancer worry scale were significantly more likely to agree that seeking reassurance, for advice about risk factors for cancer, and interest in genetic testing were important reasons. They were also more likely to agree that knowledge about the family history of cancer, a desire to help other family members,
What motivates interest in attending a familial cancer genetics clinic?
165
Table 2. Group comparisons of reasons (t-test). Variable Cancer status t
P
Having children
Cancer worry
t
t
P
–3.07
0.003**
–3.29 –2.66 –3.53 –2.39 –2.75 –2.63
0.001*** 0.009** 0.001*** 0.018* 0.007** 0.010**
–6.26
0.000***
P
Instigates referral t
P
–2.18
0.031*
–2.04
0.043*
–4.04
0.000***
Reason Family history Personal risk Genetic testing Screening Risk factors Family risk Prevention Help family Reduce risk Reassurance Advised
–3.17
0.002**
–3.65 –2.86
0.000*** 0.007**
–2.70
0.010**
–2.55
0.014*
–4.59 –2.16 –1.97
0.000*** 0.036* 0.000***
* P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001. Table 3. Group comparisons of reasons (one way ANOVA). Variable Perceived risk f
Age P
f
Level of education P
f
P
8.56
0.000***
3.74 5.77 3.78 3.15 7.07 3.98
0.026* 0.004** 0.025* 0.045* 0.001*** 0.021*
Reason Family history Personal risk Genetic testing Screening Risk factors Family risk Prevention Help family Reduce risk Reassurance Advised
3.75 4.09
4.03
0.026* 0.019* 3.28
0.013*
5.15
0.001***
0.020*
* P ≤ 0.05, ** P ≤ 0.01, *** P ≤ 0.001.
advice about cancer prevention, access to screening and establishing the risk for family members were important. Who first raised issue of family history Participants who had been the first to raise the issue of a family history of cancer (n = 86–89) were less likely to agree that advice from a GP was an important motivator. They were also less likely to agree that helping the family or advice about risk factors were important, compared to those who had not been the first to raise the family history (n = 71–73). Risk perception There were three significant group differences based on participants’ perception of personal risk (high n = 48–50, moderate n = 72–75, lower n = 34–35): access
to screening, advice about genetic testing, and to obtain reassurance (Figure 4d). Post hoc contrasts were carried out to identify the specific group differences. Each one showed a significant difference between the high and moderate risk perception groups: access to screening (f = –0.4022, P ≤ 0.05), advice about genetic testing (f = –0.3467, P ≤ 0.01) and to obtain reassurance (f = –0.5723, P ≤ 0.01). Family history There were no significant differences in the motivation of those with a predominant family history of breast, ovarian, colorectal and other types of cancer (breast n = 95–98, ovary n = 22–23, bowel n = 33, other n = 5–6), nor were there any significant findings for the reasons endorsed as a function of the total number of family members diagnosed with cancer.
166
L. Fraser et al.
Age
The proportion of referrals at different levels of estimated risk of developing cancer has not changed, despite increasing awareness of cancer genetics in the intervening period: lower risk (25% vs 26%), moderate risk (44% vs 49%), high risk (31% vs 25%). Most referrals were made by GPs or hospital doctors, frequently at the instigation of those referred. More people at lower risk were referred by GPs (33% vs 17%) which is consistent with Wonderling et al.’s original findings (29% vs 18%). Previous studies suggest that despite referral guidelines, GPs may refer lower risk patients to familial cancer clinics because of a complex interplay of professional and patient-centred reasons. These include a lack of knowledge about cancer genetics and how to estimate risk accurately, a belief in their traditional role as generalists resulting in a tension between the role they envisage for themselves and that of policy makers, and their concern to alleviate anxiety in patients wishing to be seen by a specialist [24–27]. The disparity between the proportion of men and women being referred is noteworthy, as is the tiny proportion of people from ethnic minorities, neither of which have changed appreciably since the time of Wonderling et al.’s study. That women account for a significantly greater proportion of referrals than men is not surprising. Most referrals are currently for family histories of breast cancer. Whilst men can develop breast cancer, they and their families may be unaware that they can pass on a mutated gene that predisposes towards breast (and other) cancers. The relative underuse of health services by men is well documented across all spheres of health care [28]. Women, on the other hand, are recognised as being more willing to seek medical advice and taking responsibility for the health of other family members, and this has been observed in women undergoing genetic testing and for passing on information about cancer susceptibility to other relatives [29, 30]. Factors contributing to under-use by ethnic minorities are likely to reflect complex cultural issues that interfere with service utilization, including the perceived importance of inheritance in cancer susceptibility and the incompatibility of culturally determined attitudes towards genetic counselling and genetic testing [31]. There may be general problems of communication with health care professionals, such as language and literacy barriers, compounded by a limited professional understanding of multicultural differences. More research is needed on which to base the provision of culturally appropriate services [32, 33]. The study also examined why people were interested in attending a family cancer clinic. Heightened personal risk perception has been reported in women with a family history of breast cancer [5]. However, Brain et al. acknowledge that their focus on a single main reason may have masked the contribution of other reasons that are not mutually exclusive. In order to facilitate comparison with Brain et al.’s data, we examined the main reasons in our sample. We also examined the average
There were significant group differences in the reasons endorsed as a function of participants’ age (n = 155–159). People aged between 51 and 70 years were more likely to agree that helping the family was an important reason, compared to those aged 21 to 40 years, and those aged 51 to 60 years were more likely to agree that establishing the family risk was important compared to people aged 31 to 50 years. Education There were also significant group differences as a function of level of education (n = 155): people who left school before the age of 16 years were more likely to agree that seven of the eleven reasons were important, compared to those educated beyond secondary school level. These included awareness of the family history, a desire to reduce personal risk, establishing family risk and to help the family, access to screening, advice about risk factors, and prevention. Marital status The reasons endorsed by participants as a function of marital status (married/cohabiting vs other) were not significant. It was not possible to explore the motivation of participants as a function of gender or ethnicity because the number of men and people from non-white ethnic groups were not large enough to permit comparisons.
Discussion Our findings suggest that there have been few changes in the pattern of service users or the source of referrals since they were first assessed by Wonderling et al. in their national survey of cancer genetics services in 1998 [2]. There has been little change in the proportion of the three main cancer types predominant in family histories: breast cancer (61% vs 63%, Wonderling et al. [2]), colorectal cancer (21% vs 16%), and ovarian cancer (14% vs 12%). GPs are known to be concerned about the therapeutic gap in the provision of effective interventions available for those at increased risk of developing cancer to reduce risk [21]. It is disappointing in this respect to observe the relatively low proportion of referrals for a family history of colorectal cancer, because surveillance programmes for those at high risk of colorectal cancer are both available and effective [22]. Embarrassment about bowel-related health problems may inhibit people from raising their concerns about a family history of colorectal cancer, and health campaigns have recently focused on raising public awareness and promoting discussion about colorectal cancer to address this [23].
What motivates interest in attending a familial cancer genetics clinic? ratings of agreement for each of the eleven reasons listed across a number of grouping variables, reflecting the diverse composition of our sample, in order to give a more comprehensive picture of what motivates interest. The proportion of people who were the first to raise the issue of their family history with a health professional (55% vs 60%, Brain et al. [5]) and the proportion who stated that the referral had been initiated by them (16%) appear unchanged. The main reasons for interest in attending a familial cancer clinic reported by our unaffected participants were similar to those reported by Brain et al.’s [5] (unaffected) sample. Interest in establishing personal risk of developing cancer was the most frequently endorsed reason (35% vs 29%), followed by knowledge of cancer in the family (17% vs 19%), and establishing the risk for other family members (10% vs 13%). However the two samples differ in the amount of interest expressed about gaining access to screening. Twice as many people in our sample said that screening was the main reason for wanting to attend (15% vs 7%), while interest in genetic testing (4% vs 10%) and reassurance (2% vs 11%) were lower. Examination of significant group differences in mean levels of agreement reveal that screening is rated as more important by those with children, along with those who perceive their risk of developing cancer as much higher than average, and those who were categorised as having high levels of cancer worry. Similarly, interest in genetic testing and reassurance is more likely to be endorsed as an important reason by those who perceive their risk as high and those with high levels of cancer worry. Our data provide support for Brain et al.’s finding that older people are more interested in attending in order to find out about the risk to other family members. We also replicated their finding that marital status and the number of affected relatives were not significant in differentiating people in their reasons for wanting to attend. However, unlike Brain et al. [5], we did find that participants’ level of education was a significant factor. Just over 15% of our sample had left school before the age of 16 years and this group was more likely to agree about the importance of obtaining information on a wide range of topics compared to those who had continued in education beyond that age. There was evidence of a positive correlation between the current age of unaffected participants and the age at which the family member (identified by participants as having had the greatest impact on their perception of their own cancer risk) had been diagnosed with cancer. The strength of this relationship is consistent with that reported by Brain et al. [5] and offers further support for the suggestion that the timing of seeking advice about a family history of cancer is influenced by an awareness of the similarity in age of an affected family member. The motivation of people who have previously been
167
diagnosed with cancer can be distinguished from their non-affected counterparts by their altruistic concern for other family members. Not only is this group more likely to agree that establishing the risk of developing cancer for other family members and helping the family are important reasons for wanting to attend, they also appear less likely to have been advised to attend by their GP, indicating a degree of self-motivation. Having children is also associated with altruistic behaviour: establishing family risk and helping the family are clearly more salient issues for those who have children compared to those who do not. In addition, those with children are more likely to agree that access to screening and advice about cancer prevention are important reasons for attending. This may reflect parents’ concern to safeguard their own health in order to be able to continue to care for their children. Anxiety about developing cancer is an important reason for wanting an appointment at a genetics centre/ clinic. People with higher scores on the cancer worry scale are significantly more likely to agree with most of the reasons assessed, compared to those who are less anxious. The only areas of consensus between people who are anxious and those who are less anxious relates to the level of agreement observed about the importance of establishing personal risk, seeking advice about risk reduction, and following the advice of a GP. People who believe their risk of developing cancer is high are more likely to agree that interest in obtaining information about genetic testing, access to screening and for reassurance, are important reasons for attending compared to those who perceive their risk to be moderate. Almost half of our participants were accurate in their estimates of personal risk and this proportion is consistent with previous research. However, contrary to previously reported findings of a tendency of women attending family breast cancer clinics in the UK to under-estimate their risk of developing cancer [34–36], the errors in our study sample appear to be fairly well balanced between over- and under-estimates. These findings provide useful insight into the motivation of different sub-groups of people referred to a familial cancer genetics centre/clinic. This information will be useful for GPs and other health care professionals at primary care level, at cancer units, at family cancer clinics, and for clinical geneticists and genetic counsellors seeing patients with a family history of cancer.
Acknowledgements We would like to acknowledge the support received from members of the Steering Committee for the British Family Cancer Record (Tim Bishop, John Burn, Lauren Padmore, Peter Sasieni, Eamonn Sheridan, Huw Thomas, Gwen Turner, Richard Turner, Robin Waxman) and the research co-ordinators at participating centres
168 (Lynn Mann, Julie Miller, Matthew Noyek, Helen Stroud). This research was funded by Cancer Research UK and the Community Fund.
References 01. Genetics and cancer services. Report of a Working Group for the Chief Medical Medical Officer, Department of Health. Working Group for the Chief Medical Officer, editor. London: Department of Health, 1998. 02. Wonderling D, Hopwood P, Cull A et al. A descriptive study of UK cancer genetics services: an emerging clinical response to the new genetics. Br J Cancer 2001; 85(2): 166–70. 03. Breast cancer gene queries surge. BBC News, 29 September 2001. 04. Richards MP. Genetic counselling for those with a family history of breast or ovarian cancer – current practice and ethical issues. Acta Oncol 1999; 38(5): 559–65. 05. Brain K, Gray J, Norman P et al. Why do women attend familial breast cancer clinics? J Med Genet 2000; 37(3): 197–202. 06. Collins V, Halliday J, Warren R, Williamson R. Assessment of education and counselling offered by a familial colorectal cancer clinic. Clin Genet 2000; 57(1): 48–55. 07. Hallowell N, Murton F, Statham H et al. Women’s need for information before attending genetic counselling for familial breast or ovarian cancer: a questionnaire, interview, and observational study. BMJ 1997; 314(7076): 281–3. 08. Hopwood P. Psychological issues in cancer genetics: current research and future priorities. Patient Educ Couns 1997; 32(1–2): 19–31. 09. Julian-Reynier C, Eisinger F, Chabal F et al. Cancer genetics clinics: target population and consultees’ expectations. Eur J Cancer 1996; 32A(3): 398–403. 10]. Richards M, Hallowell N, Green J et al. Counseling families with hereditary breast and ovarian cancer: a psychosocial perspective. J Genet Couns 1995; 4: 219–33. 11. Hopwood P. Breast cancer risk perception: what do we know and understand? Breast Cancer Res 2000; 2(6): 387–91. 12. Shankar A, Chapman P, Goodship J. Genetic counselling: do we recognise and meet the consultands’ agenda? J Med Genet 1999; 36(7): 580–2. 13. Nelson K, Garcia RE, Brown J et al. Do patient consent procedures affect participation rates in health services research? Med Care 2002; 40(4): 283–8. 14. Townsend P. Toward equality in health through social policy. Int J Health Serv 1981; 11(1): 63–75. 15. Townsend P, Phillimore P, Beattie A. Health and Deprivation: Inequality and the North. London: Routledge, 1988. 16. Majeed FA, Cook DG, Poloniecki J, Martin D. Using data from the 1991 census. BMJ 1995; 310(6993): 1511–4.
L. Fraser et al. 17. Lerman C, Seay J, Balshem A, Audrain J. Interest in genetic testing among first-degree relatives of breast cancer patients. Am J Med Genet 1995; 57(3): 385–92. 18. Lerman C, Trock B, Rimer BK et al. Psychological side effects of breast cancer screening. Health Psychol 1991; 10(4): 259–67. 19. Lerman C, Trock B, Rimer BK et al. Psychological and behavioral implications of abnormal mammograms. Ann Intern Med 1991; 114(8): 657–61. 20. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics 1997; 33: 159–74. 21. Elwyn G, Gray J, Iredale R. Tensions in implementing the new genetics. General practitioners in south Wales are unconvinced of their role in genetics services. BMJ 2000; 321(7255): 240–1. 22. Winawer S, Fletcher R, Rex D et al. Colorectal cancer screening and surveillance: clinical guidelines and rationale – update based on new evidence. Gastroenterology 2003; 124(2): 544–60. 23. Naked bottoms in cancer campaign. BBC News, 4 April 2002. 24. Elwyn G, Iredale R, Gray J. Reactions of GPs to a triage-controlled referral system for cancer genetics. Fam Pract 2002; 19(1): 65–71. 25. Kumar S, Gantley M, Elwyn G, Iredale R. Integrating genetics into primary care in practice. BMJ 2001; 323(7325): 1367. 26. Lucassen A, Watson E, Harcourt J et al. Guidelines for referral to a regional genetics service: GPs respond by referring more appropriate cases. Fam Pract 2001; 18(2): 135–40. 27. Watson E, Austoker J, Lucassen A. A study of GP referrals to a family cancer clinic for breast/ovarian cancer. Fam Pract 2001; 18(2): 131–4. 28. Verbrugge LM. Gender and health: an update on hypotheses and evidence. J Health Soc Behav 1985; 26(3): 156–82. 29. Green J, Richards M, Murton F et al. Family communication and genetic counseling. The case of hereditary breast and ovarian cancer. J Genet Couns 1997; 6: 45–60. 30. Hallowell N. Doing the right thing: genetic risk and responsibility. Sociol Health Illness 1999; 21(5): 597–621. 31. Meiser B, Eisenbruch M, Barlow-Stewart K et al. Cultural aspects of cancer genetics: setting a research agenda. J Med Genet 2001; 38(7): 425–9. 32. Wang VO. Multicultural genetic counseling: then, now, and in the 21st century. Am J Med Genet 2001; 106(3): 208–15. 33. Weil J. Multicultural education and genetic counseling. Clin Genet 2001; 59(3): 143–9. 34. Cull A, Anderson ED, Campbell S et al. The impact of genetic counselling about breast cancer risk on women’s risk perceptions and levels of distress. Br J Cancer 1999; 79(3–4): 501–8. 35. Evans DG, Burnell LD, Hopwood P, Howell A. Perception of risk in women with a family history of breast cancer. Br J Cancer 1993; 67(3): 612–4. 36. Weil J. Multicultural education and genetic counseling. Clin Genet 2001; 59(3): 143–9.
