Journal of Genetic Counseling

, Volume 16, Issue 2, pp 211–222 | Cite as

Risk Perception, Worry and Satisfaction Related to Genetic Counseling for Hereditary Cancer

  • Cathrine Bjorvatn
  • Geir Egil Eide
  • Berit Rokne Hanestad
  • Nina Øyen
  • Odd E. Havik
  • Anniken Carlsson
  • Gunilla Berglund
Original Research

Abstract

In this multi center study, genetic counseling for hereditary cancer was evaluated by assessing patients’ worry, perceived risk of developing cancer and satisfaction with genetic counseling. An overall aim was to identify characteristics of vulnerable patients in order to customize genetic counseling. In addition, agreement between patients’ and counselors’ scores was measured. A total of 275 Norwegian patients were consecutively recruited, and 213 completed questionnaires before and after genetic counseling. Patients’ perceived risk decreased after the genetic counseling session. There was incongruence between risk perception expressed as a percentage and in words. Patients were significantly less worried after counseling. Higher levels of worry were predicted by low instrumental satisfaction with counseling, high degree of perceived risk of developing cancer and younger age. In conclusion, counselors met the patients’ psychological needs to a satisfactory degree during counseling. However, patients did not fully understand their risk of developing cancer.

KEY WORDS:

perceived risk worry satisfaction genetic counseling. 

INTRODUCTION

Hereditary breast and ovarian cancer (HBOC) and hereditary non polyposis colorectal cancer (HNPCC) are two of the most common hereditary cancer syndromes in the western world (Garber and Offit, 2005; Umar et al., 2004). Genetic predispositions have been found in approximately 5%–10% of women with breast cancer. Two major breast cancer genes have been detected: BRCA1 (Miki et al., 1994) and BRCA2 (Wooster et al., 1995). A woman with a mutation in either one of these genes has a high lifetime risk of both breast cancer (up to 90%) and ovarian cancer (up to 60%) (Pasacreta, 2003; van Dooren et al., 2004). For HNPCC, there are three key DNA mismatch repair (MRR) genes well known to be linked to this cancer syndrome (MSH2, MLH1, MSH6) (Stormorken et al., 2005; Umar et al., 2004). The lifetime risk of colorectal cancer for a germ-line mutation carrier is approximately 90%. Women with mutations in these genes also have an increased risk of uterine cancer (particularly for mutations in MSH6). Other cancer diseases may also be associated with HNPCC (Ollikainen et al., 2005).

In Norway, all patients referred for possible hereditary cancer receive genetic counseling prior to genetic testing or enrollment in surveillance programs (Act, 2003-12-05).

``Genetic counseling is the process of helping people understand and adapt to the medical, psychological and familial implications of genetic contributions to disease. This process integrates the following:

  • Interpretation of family and medical histories to assess the chance of disease occurrence or recurrence.

  • Education about inheritance, testing, management, prevention, resources and research.

  • Counseling to promote informed choices and adaptation to the risk or condition'' (Resta et al., 2006).

During a genetic counseling session, the task of the physician or the genetic counselor is to impart advanced medical information to the patient using everyday language that lay people can understand, as well as providing psychological support. One important purpose is to give information that provides the basis for informed decisions, for example, regarding genetic testing and/or participation in surveillance programs for hereditary cancer (Trepanier et al., 2004).

In this context, the giving and receiving of risk estimates are of fundamental importance (Pearn, 1973). Mulvihill and Stadler (1996) used risk as a term to denote the probability that a woman with given risk factors and of a given age will develop breast cancer over a defined period of time. However, risk assessments and risk estimates are difficult to understand. Several studies of genetic counseling (Mulvihill and Stadler, 1996; Pasacreta, 2003) have found problems related to communication of understandable risk estimates to women with increased risk of breast cancer (Cull et al., 1999; Evans et al., 1994; Lerman et al., 1995; Watson et al., 1999). During the last 20 years, psychological research has established that the risk concept is multidimensional (Slovic, 2000). There are often two elements in the concept of risk: probability and consequence. When a person is informed about genetic risk, his or her perception and evaluation of that risk is specific to that individual.

The amount of literature on emotional reaction related to genetic counseling has increased in recent years (Watson et al., 2005). Cancer-related worry is often used as an outcome variable in these studies (Hopwood et al., 2004; Watson et al., 2005). Brain et al.(2002) report a significant reduction in cancer-related worry after genetic counseling among women with low or moderate risk of developing cancer, but not for women at high risk for cancer. However, in a meta-analysis no effect was observed in controlled trials on general anxiety or cancer-specific worry, although several prospective studies demonstrated short-term reductions in these outcomes (Braithwaite et al., 2004, p. 122). On the other hand, there may be subgroups of patients with high levels of worry, and it is important to identify possible predictors of a high level of worry after genetic counseling, so that more resources can be allocated to the most vulnerable patients.

Within the health system, increasing attention is being paid to patients’ satisfaction with their medical treatment and care (Ervin, 2006; Gold and McClung, 2006). There is a growing number of studies of patients’ satisfaction with genetic counseling (Hunfeld et al., 1999; Michie et al., 1997; Tercyak et al., 2004; Bresser et al., 2006; Charles et al., 2006; Wagner et al., 2005). Nordin et al. (2002) reports a high degree of satisfaction among Swedish patients undergoing genetic counseling. Michie et al.(1997) reported that knowledge was not associated with satisfaction with the information received in a genetic counseling session nor with the levels of concern or anxiety following the genetic counseling.

There is a gap today between knowledge of advanced medical DNA technology on the one hand, and knowledge of how patients understand and react to the information given in a genetic counseling session on the other (Pullman and Hodgkinson, 2006). More knowledge and research concerning the best ways of explaining risk, and understanding patients’ emotional reactions and perceptions of risk would help to develop better counseling methods.

The purpose of this study was to explore how patients perceive their risk of developing cancer. Furthermore, we investigated the patients’ level of worry. Finally, we examined the patients+ level of satisfaction, all in relation to the genetic counseling session.

METHODS

Patient data were collected immediately before and after the genetic counseling session, while the geneticist or genetic counselor (later referred to as the counselor) filled in their questionnaires immediately after the genetic counseling session.

Study Population

Patients

Patients were recruited from the genetic outpatient clinic at three university hospitals in Norway (Bergen, Trondheim and Stavanger) during 2003. The criteria for inclusion were: aged above 18 years and having at least one first- or second-degree relative with breast/ovarian cancer, colorectal cancer or both syndromes. The patient group consisted of people being counseled for cancer risk based on documented family history. Of the 275 patients invited to participate in this study, 221 (80%) agreed to participate and 213 (77%) completed all the questionnaires. All participants gave written informed consent four weeks before genetic counseling. The study was approved by the Regional Committee for Medical Research Ethics, Western Norway and The Norwegian Social Science Data Service.

Counselors

Three medical geneticists and four genetic counselors provided the counseling sessions and filled in a short questionnaire with respect to every counseled patient. The counselors’ response rate to the questionnaires was 93%.

The Patient Questionnaire

Sociodemographic and Medical Variables

Before the genetic counseling, the cancer syndromes in the family were identified. In addition, information was collected about age, gender, marital status, parity, educational level and occupational status. Existing cancer diagnoses were also recorded.

Risk Perception

Risk perception refers to the individual+s estimation of his or her absolute risk and relative risk (Slovic, 2000). Before and after the counseling session, risk perception for developing cancer was assessed by three questions based on Evans et al.(1994). Firstly, patients rated their perceived risk of developing cancer as a percentage (0–100%). Secondly, patients rated their own risk of cancer in words, using a six-point category rating scale (unlikely=1, no doubt=6). Finally, patients compared their own risk to that of other persons of the same age and gender (peers), on a five-point scale (much lower=1, much higher=5).

Level of Worry in Connection with Genetic Counseling

Patients’ worry was assessed on a seven point single item scale (no worry at all=1, worst possible worry=7), before and after genetic counseling. The scale was developed by a Swedish research-group (Nordin et al., 2002; Liden et al., 2003) and translated into Norwegian according to standard rules (forwards, backwards translated).

The worry question is designed to measure a global state specific worry in connection with the genetic counseling session. The Worry scale was moderately correlated with the anxiety part of the Hospital Anxiety and Depression scale (HAD) (r=0.35, p < 0.001), but not significant related with the depression part of HAD (r=0.08, p=0.25).

Satisfaction with Genetic Counseling

Patients completed the ``Satisfaction with Genetic Counseling Scale" (SCS) after the counseling session (Shiloh et al., 1990). This scale comprises three subscales, each with three items. The three dimensions of satisfaction (range: 3–12) were: Instrumental satisfactions (e.g., ``Did the doctor explain the condition clearly?"), Affective satisfactions (e.g., ``Did the counselor listen to what you had to say?") and Procedural satisfactions (e.g., ``How did you rate the length of time you waited from your first contact with the clinic until your first visit?"). Three individual items assessed satisfaction specifically with information, whether expectations were fulfilled, and overall satisfaction with the counseling (Nordin et al., 2002; Shiloh et al., 1990). The response format was from ``not at all satisfied"=1, to "as satisfied as possible"=4 for all items. Finally, the patients were asked if they had been offered enrollment in the surveillance program (yes, no).

Reliability of the SCS, as estimated by Cronbach+s alpha, was satisfactory (r=0.80), while Cronbach+s alpha for each of the subscales ranged from 0.54 (Procedural) to 0.70 (Affective). Shiloh et al.(1990) found similar alpha values for the whole scale (r=0.78) and for the subscales from 0.59 (Procedural) to 0.68 (Instrumental).

The Counselors’ Questionnaire

The genetic counselor questionnaire was developed by a Swedish research group (Liden et al., 2003; Nordin et al., 2002) and was based on the work of Evans et al.(1994). The questionnaire was translated into Norwegian according to standard rules. Counselors completed a questionnaire after each counseling session, recording the risk estimate s/he had communicated to the patient, the patient+s risk of developing cancer and the patient+s level of worry, all rated in the same response format as the corresponding items in the patient questionnaire. The counselors did not fill in the satisfaction with the genetic counseling scale, because this scale is based on the patient perception. They, however, responded to one item asking how satisfied they were with the counseling session (seven-point response scale). In addition, the counselor noted whether the patients had been offered inclusion in a surveillance program. Counselors’ assessment of risk evaluation and inclusion in the surveillance program were considered as the ``correct" answers.

Description of the Genetic Counseling Protocol

All patients were counseled according to the same Genetic Counseling Protocol. The genetic counseling session started with the patients+ report of their family history of cancer, in order to verify information collected from different sources during genetic investigation. Counselors’ will initially try to get an impression of the patients’ emotions regarding the cancer disease in the family.

The criteria used to include patients in surveillance program for breast and ovarian cancer (see box1) was developed by Biomed2 Demonstration Program and published in 1999 (Moller, Reis et al., 1999). The program group consisted of eleven collaborating European clinical genetic centers and their joined forces on the management of inherited breast cancer. Lifetime risk for developing breast cancer according to these criteria is calculated to be 20% or higher (Moller et al., 2002).

Information was then given about how heredity contributes to cancer diseases, and the risk of recurrence in the family.

A risk estimation was provided to the patients, detailing the lifetime risks for developing cancer in words. Lifetime risk of developing cancer in mutation carriers was given in percentage. However, none of the respondents were at this time aware of their carrier status. Patients were then given information about cancer incidence (both for sporadic and for hereditary forms), and on the development of cancer. Finally, information about surveillance programs was provided, including why, when, where and how often they would be performed, along with recommendations about self-examinations and genetic testing.

Box I. Inclusion in surveillance program

For breast ovarian cancer

Being first-degree relative (or second-degree through men) to following family history:
  • two or more relatives with early-onset breast cancer (<50 years)

  • multiple cases of breast or ovarian cancer compatible with dominant inheritance

  • a combination of early-onset breast cancer and ovarian cancer

  • demonstrated carrier of BRCA1 or BRCA2 mutations (Moller, Evans et al., 1999; NBCG, 2003).

For colorectal cancer/endometrial cancer
  • three relatives with cancer, one of whom is a first-degree relative of the other two; involving at least two generations; one or more cases diagnosed before the age of 50.

  • very small families can be considered HNPCC even if only two cancers in first-degree relatives; cancer must involve at least two generations and one or more cases must be diagnosed under age 55 (Rodriguez-Bigas et al., 1997; Vasen et al., 1999).

Statistical Methods

Descriptive analyses were used for sociodemographic and medical variables reporting mean values, standard deviation (SD) and range. The paired-sample t-test was used to compare means with repeated measures. In cases of statistical dependency between categorical variables, e.g., risk estimation from the same individuals at different occasions, McNemar-Bowker+s test for symmetry was used.

To defining patients who under-, correctly or overestimated their risk of developing cancer, we requested a discrepancy of at least two categories on the six-point scale. When defining differences between patients and counselors regarding patients’ risk compared with others of the same age and gender, we collapsed the original response options for the risk questions from five to three. This was done in order to increase response frequency at all levels and to make a clearer distinction between the levels. For example, ``When you compare your own risk of developing cancer with someone else of your age and gender, what risk do you have?": ``Much lower", ``lower", ``same", ``higher" and ``much higher". The two lower categories were collapsed into ``lower" and the two highest into ``higher". Agreement in the level of risk estimation in words between patients and counselors was quantified by estimating the linear weighted kappa.

The Kruskal-Wallis test was used to test the difference of scores between three or more independent groups when the scores were not normally distributed. In the present study, the Kruskal-Wallis test was used to investigate if there was equality in the level of worry between patients who rated their risk as low, moderate or high, respectively, according to the Claus model (1990). When comparing patients’ level of worry with the patients' self-rated level of risk, we used the cutoff point from the Claus model to define what lifetime risk should be considered low, moderate or high. The Claus model (1990) has the following cutoff points: low (<20%), moderate (20%–34%) and high (>34%) lifetime risk. However, we did not calculate or use the risk estimate from the Claus model in the individual counseling session.

Binary logistic regression analyses were used to identify predictors for high levels of worry and patients’ misunderstanding of information about the surveillance program. The logistic regression analyses were run stepwise both forwards and backwards.

All analyses were done with SPSS version 13.0 (2004), except for the weighted kappa estimation where StatXact (2004) was used. A two-tailed significance level of α =0.05 was used, and exact tests were used whenever available.

Except for analysis of variance of multiple groups, no formal corrections were made for multiple testing. Although all p-values less than 0.05 are reported, a crude adjustment for multiple testing will be to have more confidence in a result when p < 0.01.

RESULTS

In the present study, 213 out of 275 patients (77%) agreed to participate and answered the questionnaires. Among the non-respondents, 54 (20%) patients did not give their consent to inclusion while eight (3%) patients did not return the questionnaires. There were no significant differences between non-respondents and respondents with respect to age, gender or cancer diagnosis in family.

Sociodemographic and Medical Variables

The majority of the sample was female, with a mean age of 42 years (range: 18–80 years), two-thirds were married and the rate of employment was high. Half of the sample was self-referred. Among the responders, only 9% had a cancer disease (Table I), but 73% had a first-degree relative with breast/ovarian cancer or colorectal cancer. One hundred and three patients were offered inclusion in a surveillance program. The remaining patients were either males in families with breast and ovarian cancer, or too young (both male and female) to be included in a surveillance program (under 25 years for colorectal cancer, under 30 for breast/ovarian cancer). Eighty-eight percent of the respondents expected the surveillance program to be beneficial for them before the counseling session, rising to 92% after the session.
Table I.

Sociodemographic and Medical Characteristic of the Respondents (N=213)

 

n

(%)

Gender

  

 Men

35

(16.4)

 Women

178

(83.6)

Marital Status

  

 Married

138

(68.3)

 Cohabiting

25

(12.4)

 Living alone

22

(10.8)

 Divorced/separated /living with children

10

(4.9)

 Widowed

2

(1.0)

 Living with parents

5

(2.5)

 Missing

11

 

Educational level

  

 Primary school

29

(14.5)

 High school

104

(52.3)

 University

66

(33.2)

 Missing

14

 

Occupational status

  

 Full time

122

(61.0)

 Part time

49

(24.5)

 Unemployed

3

(1.5)

 Other

26

(13.0)

 Missing

13

 

Referral

  

 By doctor

107

(50.5)

 Self-referred

105

(49.5)

 Missing

1

 

Children

  

 Yes

160

(80.8)

 No

38

(19.2)

 Missing

15

 

Cancer syndrome in family

  

 Breast- ovarian

138

(64.8)

 Colorectal

49

(23.0)

 Both

26

(12.2)

Cancer diseases

  

 Yes

19

(9.2)

 No

187

(90.8)

 Missing

7

 

Perceived Cancer Risk

Perceived cancer risk addressed the following research questions: How do patients perceive their risk of developing cancer before and after genetic counseling? What is the relationship between patients’ and counselors’ ratings of patients’ risk? What is the level of agreement between risk estimated in words and as a percentage? And finally are the patients able to recall the information they were given about surveillance programs?

Before the genetic counseling, the mean of patients’ estimates of their own risk of developing cancer was 44% (SD: 20.6%). After counseling, the mean was significantly reduced to 39% (SD: 21.6 %) (paired t-test p < 0.0001).

In addition, the patients rated their own risk of cancer significantly lower on the six-point verbal scale after the counseling than before (p=0.03) (Table II).

When patients were asked to compare their own risk of developing cancer with that of someone of their own age and gender, no significant difference was found between scores before and after genetic counseling (McNemar-Bowker+s test, p=0.41).

Patients’ ranges of risk estimates for cancer in percentages were compared with each category of the patients’ estimation in words. The ranges within each category were found to be wide and overlapping before counseling, and after counseling the ranges of risk were only slightly narrowed (Table III).
Table II.

Patients’ Perception of risk of Cancer Before and After Genetic Counseling

 

After counseling

Before counseling

Unlikely

Doubtful

Uncertain

Possible

Likely

No doubt

Total

 Unlikely

1

0

0

0

0

0

1

 Doubtful

0

6

1

0

0

0

7

 Uncertain

3

4

16

20

0

0

43

 Possible

0

2

17

88

8

0

115

 Likely

0

0

6

13

15

0

34

 No doubt

0

0

0

0

0

1

1

 Total

4

12

40

121

23

1

N=201

McNemar-Bowker+s test: p=0.03

Table III.

Patients’ Range of Risk Estimation in Percentage Related to Word Categories

Before counseling

After counseling

Words

Percentage range

n

Mean

SD

Words

Percentage range

n

Mean

SD

Unlikely

0–0

1

0.0

0.0

Unlikely

0–10

4

2.5

5.0

Doubtful

0–50

7

13.9

18.2

Doubtful

0–50

12

14.7

14.6

Uncertain

0–60

40

36.4

16.9

Uncertain

0–60

38

28.3

18.1

Possibly

5–95

109

43.0

17.3

Possibly

2–75

117

40.9

16.9

Likely

30–90

34

64.8

14.9

Likely

50–90

23

66.5

14.8

No doubt

90–90

1

90.0

0.0

No doubt

90–90

1

90.0

0.0

Total

0–95

192

44.5

20.5

Total

0–90

195

39.3

21.5

Patients’ verbal evaluations of their own risk of developing cancer were compared with the counselors’ evaluations. Using a McNemar-Bowker+s test, no significant difference was discovered (p=0.22). This indicated symmetry between patients’ and counselors’ responses. However, agreement was poor with a linear weighted kappa of 0.13 (95% CI: 0.02, 0.24).

Patients and counselors showed significantly different evaluations of the patients’ risk of developing cancer compared with others of the same age and gender. The counselors more frequently assessed patients’ risk as higher compared with their peers (same age and gender), than the patients did themselves (McNemar-Bowker+s test, p < 0.0001). The agreement was poor, with a linear weighted kappa of 0.16 (95% CI: 0.07, 0.25). There was a significant change in the number of patients who over-, correctly-, or underestimated their risk before and after genetic counseling, (Table IV). More patients estimated their risk correctly after the counseling (p=0.01).

Both patients and counselors were asked if the patients had been offered inclusion in a surveillance program. A total of 27% of the patients gave wrong answers compared with the counselors. Results from a logistic regression analysis showed that higher educational levels and higher perceived risk for developing cancer were significant predictors for giving correct answers regarding inclusion in a surveillance program. We found no significant prediction value between patients giving correct answers about the surveillance program and gender, age or level of worry (Table V).

Level of Worry in Connection with the Genetic Counseling

To explore patients’ level of worry we investigated the following research questions: Do patients’ degree of worry change from before to after genetic counseling? What is the level of correspondence between patients’ own perception of worry and counselors’ perception of patients' worry? What are the predictors of post counseling worry?

Most patients reported a quite low degree of worry before the counseling (mean: 2.2, SD: 1.3, range 1–7). After the genetic counseling the mean score was even lower (mean: 1.8, SD: 1.1, t-test p < 0.0001). Before counseling, patients rated their own risk as not significantly related to worry (Kruskal-Wallis test, p=0.09). In contrast, after the counseling session, patients who rated themselves on the lowest risk level were significantly less worried (Kruskal-Wallis test, p=0.005) compared with those who perceived moderate or high risk.

A significant difference was found when comparing patients perception of their level of worry before counseling (mean score: 2.2, median score: 2) with that of the counselors’ (mean score: 3.6, median score: 3) (t-test p < 0.0001), indicating that counselors assessed patients’ level of worry higher than patients themselves did.

In a logistic regression analysis, a high level of worry was associated with lower instrumental satisfaction and higher degree of perceived risk of developing cancer. The results also showed that younger people were more likely to have a higher level of worry (Table VI).

Satisfaction with Genetic Counseling

This addressed how satisfied the patients were with their genetic counseling session.

Patients were highly satisfied with the genetic counseling session. The mean values for the three dimensions of satisfaction were: Instrumental 10.8 (SD: 1.3), Affective 11.7 (SD: 0.7) and Procedural 11.1 (SD: 1.3) (range 3–12). For the three single items in the satisfaction scale the means were 3.82 (SD: 0.4) for “expectations fulfilled,” 3.88 (SD: 0.52) for ``satisfaction with information" and 3.81 (SD: 0.4) for ``overall satisfaction" (maximum score: 4).

DISCUSSION

Patients’ perceived risk decreased after the genetic counseling session regardless of whether the risk was expressed as a percentage or in one of six verbal categories. As many as 41% of the patients believed that they had the same, or less, risk of developing cancer as their peers. The data also revealed that patients were significantly less worried after the genetic counseling session. Higher levels of worry were predicted by low instrumental satisfaction, high degree of perceived risk of developing cancer and younger age. Although the patients were exceedingly satisfied with the genetic counseling, approximately one-quarter of the patients gave the wrong answer about inclusion in the surveillance program. This may illustrate one challenge in genetic counseling that is, helping patients to recall important information.
Table IV.

Risk Estimate Before and After Genetic Counseling

 

After

Before

Correct

Overestimate

Underestimate

Total

 Correct

146

0

7

153

 Overestimate

9

10

0

19

 Underestimate

7

0

10

17

 Total

162

10

17

N=189

McNemar-Bowker+s test: p=0.01

Table V.

Logistic Regression of Answering Correct/Incorrect on Surveillance Program

Predictor

Fully adjusted odds ratio

Likelihood ratio test p-value

95% CI

Final model* odds ratio

Likelihood ratio test p-value

95% CI

Constant

0.15

  

0.34

  

Age per 10 yrs

0.98

0.89

(0.73, 1.31)

   

Gender

      

 Men (ref)

1

     

 Women

1.23

0.68

(0.47, 3.18)

   

Education

 

0.02

  

0.02

 

 Primary school (ref)

1

  

1

  

 Secondary school

3.58

 

(1.22, 10.51)

3.56

 

(1.25, 10.15)

 University degree

5.02

 

(1.57, 16.08)

5.17

 

(1.66, 16.12)

Having Cancer

 

0.27

    

 Yes (ref)

1

     

 No

2.07

 

(0.58, 7.37)

   

Level of worry

1.10

0.60

(0.77, 1.57)

   

Perceived risk per 10%

1.26

0.01

(1.05, 1.52)

1.30

<0.01

(1.09, 1.55)

*From forwards and backwards stepwise logistic regression at selection level p < 0.05.

Table VI.

Logistic Regression Analysis of Worry

Predictor

Fully adjusted odds ratio

Likelihood ratio test p-value

95% CI

Final model* odds ratio

Likelihood ratio test p-value

95% CI

Constant

11.11

  

8.22

  

Age per 10 yrs

0.71

0.05

(0.49, 1.01)

0.71

0.05

(0.50,1.00)

Gender

 

0.26

(0.53, 8.69)

   

 Men (ref)

1

     

 Women

2.15

     

Education

 

0.78

    

 Primary school (ref)

1

     

 Secondary school

0.69

 

(0.21, 2.25)

   

 University degree

0.64

 

(0.18, 2.23)

   

Instrumental satisfaction

0.69

0.01

(0.51, 0.92)

0.72

0.02

(0.54, 0.95)

Perceived risk per 10%

1.31

0.01

(1.06, 1.61)

1.35

< 0.01

(1.10, 1.65)

*From forwards and backwards stepwise logistic regression at selection level p < 0.05.

Perceived Cancer Risk

Patients’ risk categories expressed in words were compared with their understanding of risk expressed in percentages, and there was a large and overlapping range in percentage for each word category. For example, before counseling the estimations of ``50% perceived lifetime risk" was associated with all the verbal risk categories from ``unlikely" to "no doubt". Thus, it is difficult to know how the patients understood the risk concept. Personal experience may also significantly affect the way a client interprets a numerical risk (Trepanier et al., 2004). This indicates that a counselor will not really know what is meant by ``high risk", even if the patient provides the counselor with a percentage risk figure. There was also a wide range of percentage estimates within each verbal category after genetic counseling, but it was less than before. These findings reveal that patients to a large degree had the same interpretation of risk before and after genetic counseling. This is a challenge for clinical practice. One solution might be to ask the patient prior to the counseling session how they prefer to receive risk information: do they like numbers and statistics or do they prefer to discuss issues more globally in terms of elevation in their risk for developing cancer. Despite this, the reduction in both perceived risk measured in percentage and words, and the fact that significantly more patients rated a correct level of risk after counseling, indicated that only one genetic counseling session has the potential to change patients’ perception of risk.

Genetic counseling did not significantly affect patients’ risk evaluation when they compared themselves with others of the same age and gender. Half of the patients were self-referred to genetic counseling, and it was surprising that 41% of the patients estimated their risk of cancer as lower, or the same as, their peers. Counselors estimated that only 12% of patients had the same, or lower, risk than others of the same age and gender. This suggests that patients might be denying the seriousness of their situation. On the other hand, it is possible that the discrepancy was due to the patients taking into account the fact that they themselves will have surveillance program, while their peers will not. It remains an open question whether these results will have an implication for compliance with the surveillance program the patients were included in. Compliance with surveillance programs, and consequent early diagnosis and treatment of cancer, may be threatened if patients do not think that they are at a higher risk of developing cancer than other persons. This issue has been discussed in other studies particularly concerning compliance for colonoscopy (Kinney et al., 2000; Vernon, 1997).

There was a highly significant difference between patients’ and counselors' perceptions of patients’ relative risk of cancer compared with their peers. Counselors perceived these patients as a high-risk group and generally rated their risk to be higher than the normal population risk. This suggests that counselors did not succeed in stating their estimate risk to patients in such a way that the patients were able to understand. Earlier research (Liden et al., 2003) has shown that patients undergoing counseling for hereditary cancer have a tendency to overestimate the risk of cancer in the general population (58% overestimated while 7% underestimated). Other studies have revealed that women with increased risk of cancer underestimated and overestimated the population risk in more similar proportions (40% and 36%) (Watson et al., 1999). The different perspectives in risk research have given results that are not easy to compare across different studies (Brun et al., 1994).

It is important that patients with increased risk of cancer understand the information about surveillance programs. Twenty-seven percent of patients gave incorrect answers regarding inclusion in these surveillance programs following the genetic counseling session. Higher education and higher perceived risk of cancer were significant predictors of giving a correct answer regarding inclusion in the surveillance program. Higher education can be assumed to facilitate understanding, whereas counselors may be more careful to ensure patients’ understanding when dealing with patients who have a high level of risk. A higher perceived risk of cancer may also make patients more focused and concerned about what can be done to reduce risk. These findings support the use of a follow-up letter to the patient to sum up the most important issues from the counseling session. The poor recall of the surveillance program should also be a main focus in the counseling session. It is also important that counselors determine how much their patients are able to process during a counseling session, indicating that the counselor should make sure the patients understand at least the most important message provided in the counseling session. The information about surveillance program is among these important messages.

Level of Worry in Connection to the Genetic Counseling Session

Consistent with other findings (Liden et al., 2003) the average level of worry before the counseling session was very low, and was further reduced after genetic counseling. Changes of scores between 5% and 10% are noticeable by patients and are regarded as clinically significant changes (Fayers and Machin, 2000). In the present study, the reduction in the level of worry was of this magnitude, and could therefore be assumed to be clinically significant for the patients.

We found that younger age was a predictor for higher levels of worry. This is in line with earlier studies (Cull et al., 1999; Vernon et al., 1997), and was expected, since the onset of hereditary cancer occurs earlier in life than sporadic cancer. The older a person with a hereditary risk of cancer gets without developing cancer, the lower is the risk of their being a mutation carrier. Collins et al.(2000) suggest that this is likely to be related to the fear of developing cancer at an early age and the greater acceptance among older people of the fact that they may get sick. The fact that younger people slowly, but inevitably, approach the ``magic age" in the family where relatives usually get cancer, may also explain the higher level of worry younger people experience.

The study also showed that the counselors perceived patients’ level of worry significantly higher than the patients themselves reported. Similar results have been reported previously, demonstrating that health workers, such as nursing staff and physicians, overestimate patients’ level of worry (Lampic and Sjoden, 2000; Liden et al., 2003). This seems to be a general finding with proxies reporting, and can be due to the health worker+s own level of worry. These patients have often lived with the anticipation of increased cancer risk for a long time and have learned to cope with the situation, while counselors are guided by statistics and medical facts.

Satisfaction with Genetic Counseling

Patients were highly satisfied with the genetic counseling session. The mean scores of the three subscales (instrumental, affective and procedural), as well as the three single questions, were all in the very upper end of the scale range. The patients in our study were even more satisfied than those reported in earlier studies (Liden et al., 2003). This indicates that patients were satisfied with the medical knowledge the counselor had (instrumental satisfaction) and how they were cared for (affective satisfaction). In addition, they were also satisfied with the arrangement for their visits to the outpatient clinics (procedural satisfaction), for example, how long they had to wait to get an appointment. It could be assumed that counselors have met patients’ expectations and needs in a satisfying manner. Patients may also express satisfaction as a sign of gratitude, or as a desire to please. Furthermore, they may also feel relieved that counselors and the health system are willing to take care of them, to include them in a surveillance program, and to recognize their concerns.

The results of the present study indicate that satisfaction with the counselors’ medical knowledge (instrumental satisfaction) is an important factor in predicting worry after genetic counseling. A high level of instrumental satisfaction also reflects counselors’ ability to give the necessary treatment and reassurance to the patients (Shiloh et al., 1990). If patients believe that the counselor has good skills, they will probably feel safe. For some patients this feeling of safety could indeed reduce the level of worry. Previous research has shown that increased worry is associated with a greater expressed need for information in women with increased risk of breast cancer (Holloway et al., 2004). The results from the present study, together with the earlier findings suggest that highly skilled counselors are important.

Limitations of the Study

The present study has limitations that should be considered. From a strict methodological point of view, a randomized controlled trial would have been preferable, but this would be considered unethical and illegal. According to Norwegian legislation, it is illegal to perform predictive genetic investigation without giving genetic counseling (Act, 2003-12-05). The study sample was small for subgroup analyses, and consisted of consecutive patients, mostly women, seeking genetic counseling. However, the data were collected from three different hospitals in three different regions of Norway to reduce the possible influence of community characteristics.

CONCLUSIONS

In conclusion, it can be assumed that counselors met patients’ psychological needs to a satisfactory degree during the counseling session. The level of worry was reduced and the patients were highly satisfied, indicating that even one genetic counseling session has the potential to change a patient+s opinion. However, the study clearly shows that providing an estimate of risk to patients is difficult, and many patients underestimated their own risk compared with others of the same age and gender. Further research is needed to determine how patients understand and react to their risk of developing cancer. It is also necessary to develop models that increase the patients’ level of recalling important messages from the counseling session.

Notes

ACKNOWLEDGEMENTS

The authors thank all patients and counselors who participated in the study. The authors are grateful for the assistance with data collection provided by the Center of Medical Genetics and Molecular Medicine, Haukeland University Hospital, Norway and the Center of Medical Genetics, St. Olav Hospital, Norway. We also thank our funding sources: The Norwegian Research Board (Functional Genome project (FUGE) no 155757/510).

REFERENCES

  1. Act. (2003-12-05). Act regarding human medical use of biotechnology. Norwegian legislation.Google Scholar
  2. Brain, K., Norman, P., Gray, J., Rogers, C., Mansel, R., & Harper, P. (2002). A randomized trial of specialist genetic assessment: psychological impact on women at different levels of familial breast cancer risk. Br J Cancer, 86(2), 233–238.PubMedCrossRefGoogle Scholar
  3. Braithwaite, D., Emery, J., Walter, F., Prevost, A. T., & Sutton, S. (2004). Psychological impact of genetic counseling for familial cancer: a systematic review and meta-analysis. J Natl Cancer Inst, 96(2), 122–133.PubMedGoogle Scholar
  4. Bresser, P. J., Seynaeve, C., Van Gool, A. R., Brekelmans, C. T., Meijers-Heijboer, H., van Geel, A. N., et al. (2006). Satisfaction with prophylactic mastectomy and breast reconstruction in genetically predisposed women. Plast Reconstr Surg, 117(6), 1675–1682; discussion 1683–1674.Google Scholar
  5. Brun, W., Wright, G., & Ayton, P. (1994). Risk Perception:Main Issues, Approaches and Findings in Subjective Probability (pp. 295–320). Chichester: John Wiley & Sons.Google Scholar
  6. Charles, S., Kessler, L., Stopfer, J. E., Domchek, S., & Halbert, C. H. (2006). Satisfaction with genetic counseling for BRCA1 and BRCA2 mutations among African American women. Patient Educ Couns.Google Scholar
  7. Claus, E. B., Risch, N. J., & Thompson, W. D. (1990). Age at onset as an indicator of familial risk of breast cancer. Am J Epidemiol, 131(6), 961–972.PubMedGoogle Scholar
  8. Collins, V., Halliday, J., Warren, R., & Williamson, R. (2000). Cancer worries, risk perceptions and associations with interest in DNA testing and clinic satisfaction in a familial colorectal cancer clinic. Clin Genet, 58(6), 460–468.PubMedCrossRefGoogle Scholar
  9. Cull, A., Anderson, E. D., Campbell, S., Mackay, J., Smyth, E., & Steel, M. (1999). The impact of genetic counselling about breast cancer risk on women+s risk perceptions and levels of distress. Br J Cancer, 79(3–4), 501–508.PubMedCrossRefGoogle Scholar
  10. Ervin, N. E. (2006). Does patient satisfaction contribute to nursing care quality? J Nurs Adm, 36(3), 126–130.PubMedCrossRefGoogle Scholar
  11. Evans, D. G., Blair, V., Greenhalgh, R., Hopwood, P., & Howell, A. (1994). The impact of genetic counselling on risk perception in women with a family history of breast cancer. Br J Cancer, 70(5), 934–938.PubMedGoogle Scholar
  12. Fayers, M., & Machin, D. (2000). Quality of Life. Assessment, Analysis and Interpretation. West Sussex, England: Wiley.Google Scholar
  13. Garber, J. E., & Offit, K. (2005). Hereditary cancer predisposition syndromes. J Clin Oncol, 23(2), 276–292.PubMedCrossRefGoogle Scholar
  14. Gold, D. T., & McClung, B. (2006). Approaches to patient education: emphasizing the long-term value of compliance and persistence. Am J Med, 119(4 Suppl 1), S32–37.PubMedCrossRefGoogle Scholar
  15. Holloway, S., Porteous, M., Cetnarskyj, R., Anderson, E., Rush, R., Fry, A. et al. (2004). Patient satisfaction with two different models of cancer genetic services in south-east Scotland. Br J Cancer, 90(3), 582–589.PubMedCrossRefGoogle Scholar
  16. Hopwood, P. Wonderling, D., Watson, M., Cull, A., Douglas, F., Cole, T., et al. (2004). A randomised comparison of UK genetic risk counselling services for familial cancer: psychosocial outcomes. Br J Cancer, 91(5), 884–892.PubMedGoogle Scholar
  17. Hunfeld, J. A., Leurs, A., De Jong, M., Oberstein, M. L., Tibben, A., Wladimiroff, J. W, et al. (1999). Prenatal consultation after a fetal anomaly scan: Videotaped exploration of physician+s attitude and patient+s satisfaction. Prenat Diagn, 19(11), 1043–1047.PubMedCrossRefGoogle Scholar
  18. Kinney, A. Y., Choi, Y. A., DeVellis, B., Kobetz, E., Millikan, R. C., & Sandler, R. S. (2000). Interest in genetic testing among first-degree relatives of colorectal cancer patients. Am J Prev Med, 18(3), 249–252.PubMedCrossRefGoogle Scholar
  19. Lampic, C., & Sjoden, P. O. (2000). Patient and staff perceptions of cancer patients+ psychological concerns and needs. Acta Oncol, 39(1), 9–22.PubMedCrossRefGoogle Scholar
  20. Lerman, C., Lustbader, E., Rimer, B., Daly, M., Miller, S., Sands, C., et al. (1995). Effects of individualized breast cancer risk counseling: a randomized trial. J Natl Cancer Inst, 87(4), 286–292.PubMedCrossRefGoogle Scholar
  21. Liden, A., Berglund, G., Hansson, M. G., Rosenquist, R., Sjoden, P. O., & Nordin, K. (2003). Genetic counselling for cancer and risk perception. Acta Oncol, 42(7), 726–734.PubMedCrossRefGoogle Scholar
  22. Michie, S., McDonald, V., & Marteau, T. M. (1997). Genetic counselling: Information given, recall and satisfaction. Patient Educ Couns, 32(1–2), 101–106.PubMedCrossRefGoogle Scholar
  23. Miki, Y., Swensen, J., Shattuck-Eidens, D., Futreal, P. A., Harshman, K., Tavtigian, S., et al. (1994). A strong candidate for the breast and ovarian cancer susceptibility gene BRCA1. Science, 266(5182), 66–71.PubMedCrossRefGoogle Scholar
  24. Moller, P., Borg, A., Evans, D. G., Haites, N., Reis, M. M., Vasen, H, et al. (2002). Survival in prospectively ascertained familial breast cancer: analysis of a series stratified by tumour characteristics, BRCA mutations and oophorectomy. Int J Cancer, 101(6), 555–559.PubMedCrossRefGoogle Scholar
  25. Moller, P., Evans, G., Haites, N., Vasen, H., Reis, M. M., Anderson, E. et al. (1999). Guidelines for follow-up of women at high risk for inherited breast cancer: consensus statement from the Biomed 2 Demonstration Programme on Inherited Breast Cancer. Dis Markers, 15(1–3), 207–211.PubMedGoogle Scholar
  26. Moller, P., Reis, M. M., Evans, G., Vasen, H., Haites, N., Anderson, E. et al. (1999). Efficacy of early diagnosis and treatment in women with a family history of breast cancer. European Familial Breast Cancer Collaborative Group. Dis Markers, 15(1–3), 179–186.PubMedGoogle Scholar
  27. Mulvihill, J. J., & Stadler, M. P. (1996). Breast cancer risk analysis and counseling. Clin Obstet Gynecol, 39(4), 851–859.PubMedCrossRefGoogle Scholar
  28. NBCG, N. B. C. G. (2003). Brystkreft -- Diagnostikk og Behandling. Oslo:The Norwegian Cancer Society.Google Scholar
  29. Nordin, K., Liden, A., Hansson, M., Rosenquist, R., & Berglund, G. (2002). Coping style, psychological distress, risk perception, and satisfaction in subjects attending genetic counselling for hereditary cancer. J Med Genet, 39(9), 689–694.PubMedCrossRefGoogle Scholar
  30. Ollikainen, M., Abdel-Rahman, W. M., Moisio, A. L., Lindroos, A., Kariola, R., Jarvela, I, et al. (2005). Molecular analysis of familial endometrial carcinoma: A manifestation of hereditary nonpolyposis colorectal cancer or a separate syndrome? J Clin Oncol, 23(21), 4609–4616.PubMedCrossRefGoogle Scholar
  31. Pasacreta, J. V. (2003). Psychosocial issues associated with genetic testing for breast and ovarian cancer risk: an integrative review. Cancer Invest, 21(4), 588–623.PubMedCrossRefGoogle Scholar
  32. Pearn, J. H. (1973). Patients+ subjective interpretation of risks offered in genetic counselling. J Med Genet, 10(2), 129–134.PubMedCrossRefGoogle Scholar
  33. Pullman, D., & Hodgkinson, K. (2006). Genetic knowledge and moral responsibility: ambiguity at the interface of genetic research and clinical practice. Clin Genet, 69(3), 199–203.PubMedCrossRefGoogle Scholar
  34. Resta, R., Biesecker, B. B., Bennett, R. L., Blum, S., Estabrooks Hahn, S., Strecker, M. N., et al. (2006). A new definition of genetic counseling: national society of genetic counselors+ task force report. J Genet Couns, 15(2), 77–83. DOI: 10.1007/s10897-005-9014-3Google Scholar
  35. Rodriguez-Bigas, M. A., Boland, C. R., Hamilton, S. R., Henson, D. E., Jass, J. R., Khan, P. M., et al. (1997). A National Cancer Institute Workshop on Hereditary Nonpolyposis Colorectal Cancer Syndrome: meeting highlights and Bethesda guidelines. J Natl Cancer Inst, 89(23), 1758–1762.PubMedCrossRefGoogle Scholar
  36. Shiloh, S., Avdor, O., & Goodman, R. M. (1990). Satisfaction with genetic counseling: dimensions and measurement. Am J Med Genet, 37(4), 522–529.PubMedCrossRefGoogle Scholar
  37. Slovic, P. (2000). The perception of risk. London: Earthscan Publications Ltd.Google Scholar
  38. SPSS13.0. (2004). Users manual. Chicago.Google Scholar
  39. StatXact. (2004). Users manual, Cytel Statistica Software.Google Scholar
  40. Stormorken, A. T., Bowitz-Lothe, I. M., Noren, T., Kure, E., Aase, S., Wijnen, J., et al. (2005). Immunohistochemistry identifies carriers of mismatch repair gene defects causing hereditary nonpolyposis colorectal cancer. J Clin Oncol, 23(21), 4705–4712.PubMedCrossRefGoogle Scholar
  41. Tercyak, K. P., Demarco, T. A., Mars, B. D., & Peshkin, B. N. (2004). Women+s satisfaction with genetic counseling for hereditary breast-ovarian cancer: psychological aspects. Am J Med Genet A, 131(1), 36–41.PubMedCrossRefGoogle Scholar
  42. Trepanier, A., Ahrens, M., McKinnon, W., Peters, J., Stopfer, J., Grumet, S. C., et al. (2004). Genetic cancer risk assessment and counseling: recommendations of the national society of genetic counselors. J Genet Couns, 13(2), 83–114.PubMedCrossRefGoogle Scholar
  43. Umar, A., Boland, C. R., Terdiman, J. P., Syngal, S., de la Chapelle, A., Ruschoff, J., et al. (2004). Revised Bethesda Guidelines for hereditary nonpolyposis colorectal cancer (Lynch syndrome) and microsatellite instability. J Natl Cancer Inst, 96(4), 261–268.PubMedCrossRefGoogle Scholar
  44. van Dooren, S., Rijnsburger, A. J., Seynaeve, C., Duivenvoorden, H. J., Essink-Bot, M. L., Tilanus-Linthorst, M. M., et al. (2004). Psychological distress in women at increased risk for breast cancer: the role of risk perception. Eur J Cancer, 40(14), 2056–2063.PubMedCrossRefGoogle Scholar
  45. Vasen, H. F., Watson, P., Mecklin, J. P., & Lynch, H. T. (1999). New clinical criteria for hereditary nonpolyposis colorectal cancer (HNPCC, Lynch syndrome) proposed by the International Collaborative group on HNPCC. Gastroenterology, 116(6), 1453–1456.PubMedCrossRefGoogle Scholar
  46. Vernon, S. W. (1997). Participation in colorectal cancer screening: a review. J Natl Cancer Inst, 89(19), 1406–1422.PubMedCrossRefGoogle Scholar
  47. Vernon, S. W., Gritz, E. R., Peterson, S. K., Amos, C. I., Perz, C. A., Baile, W. F., et al. (1997). Correlates of psychologic distress in colorectal cancer patients undergoing genetic testing for hereditary colon cancer. Health Psychol, 16(1), 73–86.PubMedCrossRefGoogle Scholar
  48. Wagner, A., van Kessel, I., Kriege, M. G., Tops, C. M., Wijnen, J. T., Vasen, H. F., et al. (2005). Long term follow-up of HNPCC gene mutation carriers: compliance with screening and satisfaction with counseling and screening procedures. Fam Cancer, 4(4), 295–300.PubMedCrossRefGoogle Scholar
  49. Watson, M., Kash, K. M., Homewood, J., Ebbs, S., Murday, V., & Eeles, R. (2005). Does genetic counseling have any impact on management of breast cancer risk? Genet Test, 9(2), 167–174.PubMedCrossRefGoogle Scholar
  50. Watson, M., Lloyd, S., Davidson, J., Meyer, L., Eeles, R., Ebbs, S., et al. (1999). The impact of genetic counselling on risk perception and mental health in women with a family history of breast cancer. Br J Cancer, 79(5–6), 868–874.PubMedCrossRefGoogle Scholar
  51. Wooster, R., Bignell, G., Lancaster, J., Swift, S., Seal, S., Mangion, J., et al. (1995). Identification of the breast cancer susceptibility gene BRCA2. Nature, 378(6559), 789–792.PubMedCrossRefGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC 2006

Authors and Affiliations

  • Cathrine Bjorvatn
    • 1
    • 2
    • 6
  • Geir Egil Eide
    • 1
    • 3
  • Berit Rokne Hanestad
    • 1
  • Nina Øyen
    • 1
    • 2
  • Odd E. Havik
    • 5
  • Anniken Carlsson
    • 1
  • Gunilla Berglund
    • 4
    • 5
  1. 1.Department of Public Health and Primary Health CareUniversity of BergenBergenNorway
  2. 2.Center of Medical Genetics and Molecular MedicineHaukeland University HospitalBergenNorway
  3. 3.Centre for Clinical ResearchHaukeland University HospitalBergenNorway
  4. 4.Department of Public Health and Caring SciencesUppsala UniversityUppsalaSweden
  5. 5.Department of Clinical PsychologyUniversity of BergenBergenNorway
  6. 6.Cathrine Bjorvatn, Department of Public Health and Primary Health CareUniversity of BergenBergenNorway

Personalised recommendations