This study investigated the effect of a graphical presentation format in addition to a frequency format among healthy women with a family history of breast cancer attending genetic counseling. There was an overall increase in women’s accuracy of lifetime risk estimation from baseline to follow-up. However, no significant differences were found between the group who received the frequency format compared to the group who received the frequency format with a graphical display in terms of understanding of risk, psychological well-being and preventive intentions. No differences were found in women’s evaluation of the counseling process.
The BRISC study is a field study; in other words, it is a clinical trial which offers a unique opportunity to evaluate effects of different formats of communicating risks to counselees having to make real-life decisions. The study was a multicenter trial encompassing three out of nine Dutch familial cancer clinics. Women were allocated to receiving a particular risk communication format depending on the time they entered the study. Not only were the direct effects of communication assessed but also the effects at 6-month follow-up. In interpreting the results, one must, however, realize that any effect of different formats for risk communication on people’s perception and decision-making is bound to be small, in particular since this study involved a standard counseling session before the risk communication consultation (intervention) where the risks already had been discussed. Moreover, the sample was powered to detect a difference in risk accuracy. However, within each condition women were already quite accurate with regard to their own risk estimate leaving little room for improvement in both groups. Furthermore, for some women in both groups the risk status had changed as a result of genetic testing, as one would expect to happen in real life. In these cases, risk was based on the self-reported test results, which might differ from their actual risk status. For logistical reasons, actual test results were not recorded in this study. Moreover, the study measured women’s intentions to preventive behavior, and it is well known that there is often a discrepancy between intentions and actual behavior. Some caution should be taken in generalizing the results to all women with familial cancer since women with low education and women of ethnic minority groups were underrepresented. It is well known that in general women who come to the familial cancer clinics are higher educated women and therefore not representative for the whole population. Also, at 6 months, loss to follow-up was 25% and 22% of the women in the frequency format and frequency format plus graphical display group, respectively. To study differences in preventive intentions, as we intended to do for a small subgroup of women with low risk, a larger sample is probably needed.
This study showed no significant effects of an additional graphical display to a frequency format on the outcome measures. However, at 6-month follow-up, women who received the additional graphical display had higher risk perceptions, although not significantly higher, and, for women in the lowest risk category, higher intentions to have yearly mammography compared to women who received the frequency format only. There is mixed evidence regarding graphical displays improving understanding or aiding decision-making [31
]. Earlier studies did not find an effect of population array displays on improved understanding, but did have a higher affective impact, with the effect being perceived as larger compared to numerical formats [9
]. Numbers are shown to communicate more detailed or precise informational aspects, whereas graphical displays have sometimes been shown to better communicate the most significant message or gist (general impression) [15
]. It may be that in the present study, the frequency format represented as additional graphical information did not add to the accuracy of the understanding of the information already presented as numerical frequencies. Hence an individual who understands the frequency format may not need the additional graphic display to comprehend the information.
In this study, only icon arrays were evaluated, while other graphical displays such as bar charts may also be used. Waters et al. [24
], for example, demonstrated that bar graphs led to better understanding than numerical risk information only. Bar charts may be particularly helpful when comparing multiple risks [32
], and are, for example, available to support shared decision-making for women with high breast cancer risks [33
]. Ghosh et al. [22
] found that breast cancer risk communication using a graphical display (icons) accompanied by a bar graph can improve short-term accuracy among women who perceived very high risks (>50%). The question remains as to whether adding a graphical display may have adverse effects since it increases the quantity of information and possibly also the overestimation of risks.
It has been shown that inadequate perception of risk may lead to screening that is not consistent with the recommendations for their actual risk category [34
]. The present study showed that intentions regarding screening among women with a low to slightly increased breast cancer risk actually decreased after counseling, and thus were more in accordance with the guidelines, suggesting that these women understood the consequences of their risk more correctly after counseling.
Although risks are generally assumed to be important for decision-making, the results suggest that the way in which risks are presented does not influence women’s intentions, either because the presentation format has no effect on their understanding of the risks, or because women do not consider risks important for their decisions. For counselees, the risk level, in whatever form it is presented, may be less relevant compared to other factors, e.g. emotions such as worry and pre-existing beliefs [34
]. It has also been argued that personal characteristics such as cognitive ability and the ability to understand graphs (graph literacy) may influence the perception and comprehension of risks [19
]. Future studies need to consider who might or might not benefit from different formats of health risks communication, and whether certain formats may thereby overcome differences in cognitive ability.
Up till now, in most studies, the majority of unaffected women with a family history of breast cancer overestimated their breast cancer risk [4
]. In contrast, our study showed that nearly eighty percent of women in both groups accurately reported their breast cancer risk before standard genetic counseling, leaving little room for improvement and comparison between groups. The women seemed thus better informed about their own risk than women in other studies. One explanation may be the increasing attention for (hereditary) breast cancer in the media in recent years and the fact that more women and their families ask to be referred for counseling, suggesting a higher current awareness of familial risks. Nevertheless, some misunderstanding of risk prevailed, for example only one third of women could accurately report the population risk before counseling. An alternative explanation is the methodological differences between studies, caused by a wide range of risk accuracy measures [38
]. A systematic review of the impact of genetic counseling on risk perception accuracy has shown that accurate risk perception can be defined in many different ways [4
]. The authors of this review argued that risk perception accuracy should be defined as correctly counseled risk (i.e. in accordance with the clinician’s estimate). In the presented study, we have chosen to define accuracy as falling within the correct risk category, as this was how women’s own risk was actually counseled during standard genetic counseling.