Does enhanced information at cancer genetic counseling improve counselees’ knowledge, risk perception, satisfaction and negotiation of information to at-risk relatives? – a randomized study

Abstract

Purpose. The aim of the present randomized intervention study was to investigate the effect of receiving extended cancer genetic information on counselees’ knowledge, risk perception, information sharing and satisfaction with the service. Methods. In total, 147 counselees, affected by cancer and/or a family history of cancer, were randomized to extended or standard information. The levels of counselees’ knowledge and personal risk estimations were measured at four time points. In addition, counselees’ satisfaction with the counseling and sharing of the information to at-risk relatives was assessed. The intervention included meeting a specialist nurse, learning the breaking bad news method, receiving written material and video-taped counseling sessions. Results. A significant increase in the level of knowledge in participants in the “breast cancer group” regardless of the randomization was observed over time. The correct estimation of personal risk increased significantly in both groups after two weeks, but declined at the eight month follow-up. Most of the participants had informed at-risk relatives about their visit at the cancer genetic clinic. The majority of respondents in both groups were highly satisfied with the counseling. The only observed effects of the intervention were that counselees in the intervention group were significantly more satisfied with the content of the given information and with the way of informing relatives. Conclusion. Apparently, the current genetic counseling is managed properly and extended information does not seem necessary in all cases. However, some counselees need additional sessions.

The availability of genetic testing for known mutations in various cancer genes associated with hereditary cancer, and the possibility of risk estimation based on family history of cancer makes it possible for individuals to learn whether they have an increased risk of developing cancer [1]. At-risk individuals can undergo genetic counseling at specialized cancer genetic clinics to receive further information. However, the cancer genetic information and negotiation of it is complex and might be difficult to comprehend, to cope with and to transfer.

A vast body of research has examined the impact of cancer genetic counseling on counselees’ knowledge [2], risk perception [3], emotional reactions [4], communication to close relatives [5], and satisfaction with genetic counseling [6] in individuals known to be at risk of developing cancer. In general, empirical research has demonstrated that education and genetic counseling improves cognitive outcomes such as knowledge and accuracy of risk perception [7] and reduces psychological distress [8]. Despite that, understanding the actual risk remains a problem for many individuals and counselees experience difficulties in communicating the cancer genetic information to close relatives [7].

In Sweden and many other countries, the responsibility to inform relatives relies on the counselees and not on the health-care personnel, thus the majority of cancer genetic counselees feel a “moral obligation” to inform their relatives about their genetic risks [9]. Limited knowledge about genetics, poor recall of the provided information and shortage in communication skills may lead to miscommunication of cancer genetic information within families [10] as well as give adverse effects on the counselees’ mood and/or well-being.

Receiving information or informing others of a potentially life-threatening event such as a cancer diagnosis can also cause psychological distress. How to present this “bad news”, to at-risk relatives may have an important influence on counselees’ well-being and likelihood of transferring the information to their relatives. A constructive way of communicating this information stepwise to the relatives is to employ the process of Six Step Protocol for “Breaking Bad News” as described by Buckman [11].

Evidence from the discipline of education and communication suggests that patients desire written information about their condition [12] and access to such information may result in improved medical knowledge, greater satisfaction and increased well-being [13]. Furthermore, presenting the same material in various formats, and the use of printed material and video tapes and/or computer programs has been suggested in health education [14]. According to Meade and colleagues, individuals who received video-taped or written material about colon cancer had an increased medical knowledge and recalled the information more accurately [15]. According to cognitive theories information is also best learned when broken into smaller pieces and when exposure to information occurs through multiple channels [14].

Viewed in the light of this, the purpose of the present randomized study was to investigate the effect of an informational intervention on counselees’ knowledge, risk perception, communication of information to at-risk relatives and satisfaction with the service. Provided that the intervention had positive effects on counselees’ knowledge, risk perception and communication with relatives, it was assumed that the intervention may also lead to reduced psychological distress.

Materials and methods

This study was conducted at the cancer genetic clinic of Uppsala University Hospital and with approval from the Regional Ethical Review Board in Uppsala, Sweden. The consecutive recruitment of participants started in October 2003 and was completed in January 2005. Counselees were eligible to participate if they were 18 years or older; able to read, write, and speak Swedish; and did not suffer from any mental illnesses. As all eligible counselees were asked to participate no power calculation was conducted. The randomization was performed by the administrative staff of the Uppsala University Hospital, independent of the research group. Counselees received the information about their randomization prior to the genetic counseling appointment, after they had answered to the baseline questionnaires.

The design of the present study was a randomized intervention study with four measurement time-points: before the counseling session (T1), immediately after counseling (T2, only for risk perception), after two weeks (T3) and after eight months (T4). Since some participants did not answer to all of the questionnaires and some dropped out, the numbers of participants are different through T1–T4.

The counselors were kept blind to counselees’ randomization through videotaping every counseling session and through the presence of the specialist nurse during all sessions.

Compliance with the intervention was checked through a postal questionnaire after the eight months follow-up. To confirm sharing of the information, the relatives were also asked to participate in the study and answered to a questionnaire once; however, they were not included in the intervention.

Participants

In total, 163 individuals (78% of the eligible counselees) gave their informed consent and were randomized to the intervention or control group. After the first visit five counselees passed away and 11 dropped out (n = 16). A total of 147 counselees (72%) continued to participate in the study (See Figure 1).

Figure 1.  Profile of the randomization and participant flow.

The majority of the counselees were female and the median age was 56 years (range, 23–84), mainly referred to cancer genetic counseling due to breast cancer or a family history of breast cancer (Table I). Most of the decliners (88%) were female, and their median age was 59 years (range, 25–78). Medium age of those who dropped out was 65 (range, 23–83) and 82% were female.

Table I.  Socio-demographic characteristics of participants

	Intervention (n = 73)	Control (n = 74)
Gender
Male	6	8
Female	67	66
Health status
Cancer patients	28	26
Non-affected counselees	45	48
Referred due to cancer/family history of
Breast cancer	45	49
Breast/ovarian cancer	17	18
Colorectal cancer	11	7
Level of the risk estimated by the geneticist
Non-affected counselees	n (%)	n (%)
≤ 20%	8 (20)	9 (19)
21–40%	29 (72.5)	37 (77)
> 40%	2 (5)	2 (4)
Affected counselees’ relatives
≤ 20%	5 (15)	6 (23)
21–40%	23 (70)	17 (65.5)
> 40%	3 (9)	1 (4)

At the first visit, the counselees were asked for permission to contact their first degree relatives. Counselees presented 124 relatives. Information about the purpose of the study was sent to these relatives and they were asked to participate. Eighty-two relatives (66%) gave informed consent and filled out the questionnaires they had received by mail. Finally, 147 counselees (133 female and 14 male) and 82 of their relatives (57 female and 24 male) participated in the study.

Procedure

Counselees received written information about the study and a written consent sheet by mail. Study questionnaires were mailed to the counselees who agreed to participate and they were requested to bring their answers to the first session (for an overview of measurements see Figure 2). Non-affected counselees estimated their own risk of developing cancer and cancer-affected counselees estimated the risk for a specific close relative. The medical geneticists in charge completed a corresponding form concerning the risk for counselee or the specific relative chosen by the counselee.

Figure 2.  An overview of design and used questionnaires.

Before the counseling, and at the two week and the eight month follow-ups, the counselees were asked about their intention to inform their relatives about their visit at the cancer genetic clinic and about the level of their own risk for developing cancer. At the eight months follow-up, all participants were telephone interviewed with ten semi-structured questions about informing their relatives, their own feelings and their relatives’ reactions to the given information (data presented elsewhere). Once the eight month follow-up was completed the referred relatives were contacted in written.

Relatives received an invitation to participate in the study and an information sheet together with the name of the counselee who had passed the details about relatives to the research group. They were asked to contact the research team within two weeks if they were not interested in participating. The relatives questionnaire, investigated their level of anxiety and depression; whether they had received any information; and how they evaluated the extent of the given information (e.g. “Have you received any information about the content of the information your relative has received at the genetic counseling?”, “Do you think that you have received a sufficient amount of information?”).

Standard genetic counseling session

During the first counseling session a clinical geneticist provides information regarding, for instance, the differences between sporadic cancer versus hereditary cancer, basic genetics, and the risk of developing cancer as a carrier of a mutated gene. In addition, the geneticist estimates the risk for non-affected counselees’ or for affected counselees’ close relatives. The geneticist supplies information about genetic testing and surveillance programs and explains the importance of communicating this information to relevant relatives. Counselees who undergo genetic testing will attend an additional session in association with the disclosure of the test results. All counselees are offered further contact with a specialist nurse or a geneticist whenever they have questions or need additional support. During the period that the present study was conducted, the cancer genetic counseling was provided by five geneticists.

Intervention group

Directly after their visit, counselees in the intervention group received extended information in the following way:

The counselees met a specialist nurse, who had been present during the counseling session, and who went through the given information in detail once more. The nurse explained the pedigree again and asked counselees to estimate their risk and point out at-risk relatives. The nurse asked about counselees’ intention to inform their at-risk relatives and in those cases they did not plan to share the information with their relatives the nurse tried to find out the reason for this and helped them to overcome the perceived obstacles if possible. During this session, the nurse explained Buckman's model of “Breaking Bad News” [11], including the following steps: 1) Getting started, 2) Finding out what the relatives know, 3) Finding out how much they want to know, 4) Sharing the information, 5) Responding to feelings and 6) Planning and follow-up. During the three first steps the counselee deals with preparatory activities such as preparing a private physical setting, at the fourth step the news is delivered and the two last steps permit responding to the reactions and planning of a constructive follow-up. Counselees were supposed to use the model as an aid while informing at-risk relatives. In addition, the counselees received a pamphlet about basic genetic concepts as well as specific information concerning their type of hereditary cancer. Shortly after the counseling, they also received a videotape from the counseling session, a copy of their medical records and a copy of the pedigree, to use in communicating the given information to their at-risk relatives.

Control group

The control group received the standard information usually given at the cancer genetic clinic. The counselees in the control group also met the specialist nurse directly after the counseling session and were asked about their intentions to inform at risk relatives. However, they did not receive any further information or additional help in identification of important relatives. The presence of the specialist nurse together with the counselees in the control group was also necessary in order to keep the counselor (clinical geneticist) blind regarding counselees’ randomization. The video was mailed to individuals in the control group together with the last follow-up questionnaire, eight months later.

Compliance with the intervention regime

To check the compliance with the intervention, a 19 items questionnaire which assessed whether counselees had used the given material (e.g. “Have you watched the video you received after your visit at the cancer genetic clinic?”; “Have you read the brochure you received at the cancer genetic clinic?”), and how they evaluated the intervention (e.g.” Was the given information easy to understand?”), was constructed specifically for the study. Ratings were made on 10-point scales with defined endpoints (e.g. 0 = very difficult/totally disagree, 10 = very easy/totally agree). The questionnaire and a pre-paid envelope were sent to participants in the intervention group, in association with the eight month follow-up.

A total of 62 counselees (73%) completed the questionnaire regarding the use of the extended information material. The results showed that the majority of responders (n = 40) had watched the video; half of them (n = 22) had watched it at least once and together with someone else, usually their family or their partner (30–35%). Only three of the counselees claimed that the video was not good (in some cases depending on the quality of the recording). Almost all of the counselees in the intervention group (97%, n = 59) had read the pamphlets they had received and 80% thought it was good or very good. Thirty five percent (n = 21) had read the material together with someone else and 32% had lent it to somebody else. The majority of counselees (85%, n = 51) experienced the written material as very easy to understand and 90% (n = 54) evaluated it as sufficiently extensive. Thus, most of the additional information material that the intervention group (and not the control group) had received was used sufficiently.

Measures

Data presented in the present study were based on the following instruments:

Counselees’ questionnaires

Information about age, gender, marital and occupational status, education level, existing cancer diagnoses and family history of cancer was collected before the genetic counseling.

Psychological distress

The Hospital Anxiety and Depression Scale (HADS) [16], an established self-administered instrument with good psychometric properties [17], was used for assessment of anxiety and depression. In this study the Swedish version of HADS, which has been translated by Marianne Sullivan in 1986, using a forward backward procedure, and validated by Lundqvist and co-workers [18], was applied.

Risk perception

Counselees completed a risk perception form for developing cancer based on the work of Evans and colleagues [19]. Counselees were asked to estimate their risk both in percentage (for example 0–10%, 91–100%), and in comparison to the risk of other people of the same age and gender, on a five-point scale (slightly lower = 1, much higher = 5). Whenever there was a discrepancy of more than one step (above or below), compared to the risk figure given by the counselor, the risk estimation was defined as incorrect. When counselees estimated their own risk for developing cancer in comparison to another person of the same age and gender, no discrepancy compared to the counselors’ estimation was allowed. The counselors’ assessment was considered as the “correct” answer. This scale has been used in other studies by our research group. The scale included two questions about the counselees’ intention to inform their relatives about their visit at the cancer genetic clinic.

Knowledge

Knowledge about genetics was measured by a questionnaire about hereditary cancer and genetics. The respondents indicated whether a number of these statements were true or false. Most of the items were simple statements about genetics and some were specific about the cancer in question for the counselee. Therefore, the questionnaire was developed in three versions, one for breast cancer (11 items), one for breast-/ovarian cancer (13 items) and one for colorectal cancer (12 items). Scores were summed up to constitute a measure of knowledge about hereditary cancer and genetics. Counselees answered only to the questions corresponding to the diagnosis in question. The questionnaires were specifically developed for the present study by clinical geneticists and psychologist researchers with knowledge of psycho-oncology. The scale was checked for face and content validity.

Satisfaction with Genetic Counseling Scale (SCS)

SCS [20] consists of 12 items with four response choices from “not at all satisfied” = 0, to “as satisfied as possible” = 4. The scale is created to evaluate three dimensions (each subscale includes three items): “Instrumental satisfaction”, “Affective satisfaction”, and “Procedural satisfaction”. The ranges for all three dimensions are 0–9. Three individual items (scale range 0–3) assess specific satisfaction with the information (”Information”), whether expectations were fulfilled (“Expectations”), and overall satisfaction with the counseling (“Overall satisfaction”). Scores are summed up to constitute a measure of satisfaction for each dimension and for each item. In this study the Swedish version of the SCS, translated by Nordin and co-workers [21], was applied.

Satisfaction with given Oncogenetic Information Scale (SOIS)

A six items questionnaire was developed for the present study in order to specifically evaluate the counselees’ satisfaction with the given information. The respondents were asked to indicate their answers on a four-point scale ranging from “not at all satisfied” = 1 to “totally satisfied” = 4. Scores are summed up to constitute a measure of satisfaction. Examples of statements are: I have received satisfactory information about: “How diseases inherit” and “How I should inform my relatives”. The scale was checked for face and content validity. However, no other psychometric data are available. Figure 2 shows an overview of the instruments used at different time-points of the study.

Statistical analysis

Descriptive statistics were computed for socio-demographic and medical variables and satisfaction with the counseling. The HADS subscales of anxiety and depression were summarized according to the manual and correct answer scores on the knowledge test were summed up to constitute a measure of knowledge about hereditary cancer and genetics. The intervention and control group were compared using two-way ANOVA repeated measure analysis. Group differences in categorical variables were assessed by the Mann-Whitney U test, and the McNemar test was used to compare repeated measures in these variables, e.g. risk estimations. All reported p values (p ≤ 0.05 or p ≤ 0.01) were associated with two-tailed tests of significance. All analyses were carried out using SPSS statistical software (Statistical Package of the Social Science), version 14.0.

Results

Knowledge

The majority of participants in both the intervention and the control group had a high level of knowledge (Table II). Time had a significant effect on the counselees’ level of knowledge in the breast cancer group (F = 18 379, df = 2, p < 0.001), and there was no significant interaction effect (F = 1 189, df = 2, p = 0.307). In the breast cancer group, the mean level of knowledge increased significantly at both two week and eight month follow-ups compared to the baseline. However, no significant difference in knowledge was observed between the intervention and the control group in any of the “cancer diagnosis” groups or at any time-point of measurement (Table II).

Table II.  Counselees’ knowledge about hereditary cancer and genetics

					Two-way ANOVA
		Pre-counseling	2-week	8-month	Effects
Knowledge	n^a	Mean (CI^b)	Mean (CI)	Mean (CI)	Group	Time	Group x Time
“Breast cancer group” (Scale range 0–11)
Intervention	35	8.6 (8.29, 9.02)	9.6 (9.45, 10.06)	9.5 (8.98, 9.88)
Control	38	8.5 (8.09, 8.83)	9.1 (8.49, 9.39)	9.3 (8.83, 9.63)	N.S	p < 0.001	N.S
“Breast/ovarian cancer group” (Scale range 0–13)
Intervention	16	9.9 (9.23, 10.52)	10.7 (9.92, 11,46)	10.9(9.99, 11.76)
Control	16	10.2 (9.48, 11.02)	10.5(10.06, 10.93)	10.4 (9.70, 11,04)	N.S	N.S	N.S
“Colorectal cancer group” (Scale range 0–12)
Intervention	7	9.7 (8.05, 11.38)	10.1(8.59, 11.69)	10.1 (7.85, 12.43)
Control	5	10.6 (9.18, 12.01)	9.0 (6.68, 11.32)	10.0 (7.68, 12.32)	N.S	N.S	N.S

^a Data is missing for some counselees.

^b Confidence Interval 95%.

Perceived risk of cancer, presented as percentage

Non-affected counselees

The number of the counselees who estimated their personal risk correct at the two week follow-up showed a significant increase to 82% in the intervention group (McNemar test, p ≤ 0.005) and to 92% in the control group (McNemar test, p ≤ 0.001), but was reduced to approximately 60% in both groups at the eight month follow-up.The only observed significant difference between the intervention and the control group was immediately after the counseling, where a more accurate estimation was shown in the control group (Mann-Whitney test, p ≤ 0.01, Table III).

Table III.  Counselees’ risk estimation expressed in percentage

	Intervention		Control
	Total n	n (%)	Total n	n (%)
Number of non-affected counselees who estimated their own risk correctly
Pre-counseling	39	18 (47)^a	47	22 (47)^a
Post-counseling	38	21 (55) *	47	42 (89) ^*,^a
Two-week	34	28 (82)^a	38	35 (92)^a
Eight-month	35	21 (60)	36	22 (61)
Number of affected counselees who estimated their relatives’ risk correctly
Pre-counseling	34	10 (29)^a	29	9 (35)^a
Post-counseling	33	22 (67)^a	26	12 (46)
Two-week	30	22 (73)^a	29	17 (59)
Eight-month	26	19 (73)^a	26	19 (73)^a

^aSignificant difference within subjects over time.

*Significant difference between groups.

Affected counselees

Immediately after the counseling and at the two week follow-up, the number of individuals in the intervention group with correct risk estimation demonstrated a significant increase (67% and 73%, respectively) compared to the baseline (McNemar test, p ≤ 0.001). At the eight month follow-up, the number of the counselees who predicted the risk of their relatives’ correctly showed a significant increase (73%) compared to the baseline in both groups (McNemar test, p ≤ 0.01). No significant difference was observed between the intervention and the control group at any time-point of the measurement (Table III).

Comparative risk perception

Non-affected counselees

The percentage of the individuals who estimated their own risk compared with a person of the same age and gender (lower, the same, a little higher, higher, and much higher) correctly had significantly risen to 39% immediately after the counseling (McNemar test, p ≤ 0.05) in the intervention group. In the control group, the number of counselees who predicted their own risk correctly increased significantly to 58% at the two week follow-up (p ≤ 0.01), but declined to 20% at eight month follow-up. In addition, at the two week follow-up, a significant difference was observed between the control and intervention group (Mann-Whitney U test, p ≤ 0.05) (Table IV).

Table IV.  Counselees’ risk estimation compared to their peers

	Intervention		Control
	Total n	n^1(%)	Total n	n (%)
Number of non-affected counselees who correctly estimated their own risk compared with others
Pre-counseling	41	8 (19.5)^a	46	13 (28)^a
Post-counseling	41	16 (39)^a	46	16 (35)
Two-week	37	13 (35)*	38	22 (58)^*,^a
Eight-month	35	11 (31)	35	7 (20)
Number of affected counselees who estimated the risk of a specific relative correctly compared with others
Pre-counseling	33	8 (24)	23	7 (30)
Post-counseling	34	14 (42)	25	11 (48)
Two-week	29	11 (39)	26	10 (43.5)
Eight-month	25	5 (20)	22	4 (20)

¹Data is missing for some participants.

* Significant difference between groups.

^a Significant difference within subjects over time.

Affected counselees

No significant difference was observed between the intervention and the control group, neither within the subjects in any group nor at any time-point of the measurement, in estimating their relatives’ risk for developing cancer compared with an individual of the same age and gender (Table IV).

Satisfaction with genetic counseling measured by SCS

The majority of the counselees (77% in the intervention and 73% in the control group) were highly or totally satisfied with their visit at the cancer genetic clinic and almost all (99% in the intervention and 97% in the control group) were satisfied with the content of information given at genetic counseling (data not shown).

No significant difference was observed between the intervention and the control group in any of the subscales or single items with the exception of the participants in the intervention group who were significantly more satisfied with the content of the information given at genetic counseling (Table V).

Table V.  Counselees’ satisfaction with genetic counseling

	Intervention(n = 73)		Control (n = 73)^a
Satisfaction with Genetic Counseling Scales (SCS)	Mean	Sd	Mean	Sd	t
Dimensions(Min–Max =0–9)
Instrumental satisfaction	7.07	1.48	6.77	1.60	−1.18
Affective satisfaction	8.33	1.13	8.40	0.93	0.43
Procedural satisfaction	8.36	1.00	8.25	0.92	−0.69
Items(Min–Max =0–3)
Fulfillment of expectation	2.78	0.45	2.67	0.53	−1.35
Content of information	2.78	0.42	2.62	0.59	−1.94*
Direct level of satisfaction	2.84	0.37	2.83	0.44	−0.03
Satisfaction with Onco-genetic Information Scale (SOIS) Scales (1 = low 4 = high)		Intervention (n = 73)		Control (n = 72)
I have received satisfactory information about
How disease inherits	3.8	0.46	3.6	0.57	-1.53
My risk for developing …	3.7	0.51	3.5	0.63	-1.86
Which relatives are involved with the information I have received	3.8	0.56	3.7	0.62	-1.12
How shall I inform my relatives	3.6	0.72	3.1	0.94	-3.63**
Which preventive measures should I take	3.5	0.72	3.3	0.84	-1.54
Where shall I turn to for receiving the information I need	3.9	0.37	3.8	0.56	-0.72

^a Data is missing for some participants

*p <0.05

** p <0.01

Satisfaction with information given at genetic counseling measured by SOIS

The mean scores of all SOIS items were very high. No significant differences were found between counselees in the intervention and control group, with the exception of the item “How to inform relatives”, indicating a significantly higher level of satisfaction in the intervention group (M_interv=3.60, M_control=3.08) (Table V).

Informing relatives

Before the counseling session, most of the counselees reported that they intended to inform their at-risk relatives about the genetic counseling session (95% in the intervention and 91% in the control group) and about their risk level (100% in the intervention and 98% in the control group). At the eight month follow-up, 73% of the counselees in both groups reported that they had informed all their relatives about their visit at the cancer genetic clinic.

Furthermore, when asked at the eight month follow-up, the majority of relatives (95.5% relatives of the participants in the intervention and 89% of the relatives of counselees in the control group) reported that they were informed about counselees attending cancer genetic counseling, and about the content of given information. In addition, the majority of relatives (75% in the intervention and 67% in the control group) reported that they received a sufficient level of information. No significant differences were found between the relatives of the participants in the intervention and the control group regarding assessed variables (data not shown).

Psychological distress

Overall, the level of the anxiety and depression had decreased in both the intervention and the control group (Table VI). Time also had a significant effect on anxiety and depression in both groups (F_anxiety=5.052, df = 1,736, p ≤ 0.01; F_depression=4.134, df = 1.725, p ≤ 0.01). The level of anxiety decreased significantly at the eight month follow-up compared to the baseline, and the level of depression had decreased significantly at both the two week and the eight month follow-ups compared to the baseline. However, there was no significant difference in anxiety and depression between the intervention and the control group (p = 0.463) at any time-point of the study (Table VI).

Table VI.  The result of anxiety and depression

				Two-way ANOVA
	Pre-counseling	2-week	8-month	Effects
N = 126^a	Mean (CI^b)	Mean (CI^b)	Mean (CI^b)	Group	Time	Group X Time
Anxiety
Intervention	6.3 (5.21, 7.33)	5.7 (4.64, 6.76)	5.3 (2.28, 6.26)
Control	5.6 (4.57, 6.57)	5.4 (4.30, 6.49)	4.8 (3.78, 5.84)	N.S	P < 0.01	N.S
Depression
Intervention	3.1 (2.32, 3.83)	2.6 (1.97, 3.23)	2.7 (1.95, 3.38)
Control	3.4 (2.56, 4.17)	2.9 (2.19, 3.72)	2.5 (1.86, 3.24)	N.S	P < 0.01	N.S

^aData is missing for some counselees.

^bConfidence Interval 95%.

Discussion

According to the results of the present study, the only observed effect of the intervention was more satisfaction with the content of the given information and the way of informing relatives in the intervention group. Overall, an increased level of knowledge was observed in the participants over time, independent of the intervention. Furthermore, the number of counselees who estimated their risk correctly had increased significantly over time, however more accurate risk estimation was observed in the control group immediately after the counseling and at the two week follow-up. Participants experienced a high level of satisfaction with the received genetic counseling and the majority of counselees intended to inform their relatives about their visit at the cancer genetic clinic and about the content of the given information. Based on the contacted relatives’ reports, the majority of the counselees had indeed informed them and shared their perceived risk for developing cancer. The level of distress and anxiety among counselees was reduced significantly over time irrespective of the amount of the information they received (enhanced or standard information).

As expected, individuals with a family history of cancer appear to be aware of their susceptibility, and the results of the current study showed a high level of knowledge about genetics and hereditary cancer among participants prior to the counseling session. According to the results, the high level of knowledge persisted over time, in particular in the “breast cancer group”, where a significant increase was observed. At the same time, the small sample size of the “breast-/ovarian” and “colorectal cancer” group could be the reason why no significant increase in mean level of knowledge was observed in these groups as was noted in the “breast cancer group”.

The results of the present study, in line with other studies [22], underscore the difficulty of communicating the accurate risk of developing cancer to the counselees. It appeared most difficult for counselees to compare their own/-a specific relatives’ risk with others of the same age and gender, which also did not improve notably after the counseling (20–31% correct estimation at the eight month follow-up). A feasible explanation for these findings could be that distress and worry associated with the risk for cancer and attending genetic counseling [23] may be an obstacle for processing of the given information and comprehending the accurate risk for developing cancer [24]. Furthermore, as the information processing theory suggests, true learning arise only once information has been stored in the long-term memory and has been successfully retrieved [25]. Thus, a longer period of time may be needed for perceiving ones level of risk.

According to the education and communication disciplines, receiving information through multiple channels and in various formats leads to improved recall of the information. Surprisingly, our results demonstrate more accurate risk estimation in the control group, who had not received extended information, immediately after the counseling and at the two week follow-up, compared to the intervention group. When information is provided from various sources, it is probably quite important that exactly the same information is communicated. The fact that counselees in the intervention group received the information through different sources (the counselor and the specialist nurse) could be a reason for confusion and therefore more inaccurate risk estimation. It is possible that the counselor and the nurse have had different ways in expressing the same information or different images in their communication.

Thus, optimization of counselees’ risk perception remains a challenge in clinical practice. One way of upholding the risk estimation may be to offer additional counseling sessions a few month after the initial counseling. That notwithstanding, essential questions may be asked in this context such as “Is it indeed important that counselees understand their risk accurately?”; “Is understanding of the risk essential for counselees well-being?”; “Does accurate risk perception affect the probability of communication to the relatives?”; “Does good knowledge about genetics leads to more efficient risk communication to the relatives?”. These questions remain to be studied in prospective studies.

Although the results of the present study, in line with earlier studies [22], show a high level of satisfaction with the genetic counseling in both the intervention and control group, the finding that counselees in the intervention group were more satisfied with the content of the given information and with the way of informing relatives could be considered as an effect of the intervention. Furthermore, it could be assumed that the counselors have met the patients’ expectations and needs in a satisfying manner. The high level of instrumental satisfaction reflects the counselors’ capacity to give the necessary information about treatment and reassurance to the patients.

Our data, in agreement with another report [26], also revealed that most counselees wished to inform their relatives about the family's cancer risk. The results suggest that, irrespective of the amount of information the counselees received (enhanced or standard information), they wished to transfer the information they received at the genetic counseling to their at-risk relatives. Of course, one possible explanation could be counselees’ awareness of the fact that the researcher would contact the relatives they had mentioned earlier.

In summary, the poor effect of the intervention on outcome measures could besides confusion in receiving information about risk from two different sources depend on the natural difficulties associated with risk perception and on the ceiling effect concerning knowledge and intention to inform relatives, regarding initial high levels in both variables.

Methodological considerations

The prospective, randomized design, a long-term (eight month) follow-up, checking on compliance with the intervention and involving at-risk relatives (in order to check the reported data regarding communication of the information), are some of the important strengths of the present study.

Furthermore, the results are based on a study sample with a wide range of age (19–80) and a high response rate (78%) from a substantial geographic area in central Sweden. Therefore, the data might be generalized to the population who attend cancer genetic services throughout the country. In addition, satisfaction with genetic counseling and psychological distress were assessed with established self-administered instruments with good psychometric properties.

However, the study is not without limitations and the results need to be interpreted in the light of these potential short-comings. Probably, counselees with different diagnosis and/or different family history need different kind of information and it is possible that adjustment of the information to the counselees’ need could improve the effect of the intervention. Finally, the quality of the video tapes and access to them has not been satisfactory in a few cases.

Conclusion and research recommendations

In view of the fact that counselees in the intervention group were significantly more satisfied with the content of the given information including how to inform relatives, a possible conclusion could be that the extended information was properly managed. It seems that instead of focusing on enhanced information for all counselees, identifying individuals who are extremely worried and need more support, is more sufficient.

Considering the findings indicating that the level of distress in counselees was reduced over time, one plausible conclusion is that genetic counseling probably helps counselees to cope with their cancer or cancer susceptibility over the long term.

The findings further suggest that the way genetic counseling is being carried out today is satisfactory. However, given the fact that risk perception was a difficult issue which improved initially but deteriorated at the eight month follow-up, we suggest that genetics professionals actively offer additional counseling sessions, within six month post-counseling and that they make these services readily accessible.

Finally, intervention studies should be conducted in order to confirm these findings, and to determine which information modalities reach cancer genetic counselees most effectively and how we can help counselees to recall important information more properly.

Acknowledgements

This study would not be possible without contribution of individuals participating in the study and the financial support from the Swedish Cancer Society.

We would like to express our gratitude to Karin Eriksson, Annika Lidén and Kristina Thorsén for practical management of the study. Declaration of interest: The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.