Presenting health risk information in different formats: The effect on participants’ cognitive and emotional evaluation and decisions

https://doi.org/10.1016/j.pec.2008.07.013Get rights and content

Abstract

Objective

Effective communication of health risks plays an important role in enabling patients to make adequate decisions. There is little – though contradictory – evidence to indicate which format is most effective for communicating risks, and which risk format is preferred by counselees.

Methods

In an experiment, subjects were presented health scenarios and risk information in different formats (percentages, frequencies, and population figures) and asked to evaluate the risks and make a decision based on these.

Results

Different risk formats had different effects on respondents’ evaluation of the health risks presented. Contrary to our expectation, population figures were not evaluated as being the easiest format for all decision problems. Population figures were shown to have the biggest affective impact, and risks presented as population figures were also evaluated as significantly greater than the risks presented in other formats. The format of the presented risks influenced their decision in only one out of four decision-making situations, although in a second situation there was a similar trend.

Conclusion

This study suggests that the risk format plays a role in the decision-making process, although it remains unclear which format is the most effective in terms of understanding.

Practice implications

More experimental studies based on a theoretical analysis of the factors that promote effective risk communication are needed in the general population as well as in clinical settings with patients actually experiencing the risks and making the decisions.

Introduction

The role of individuals in the decision-making process with regard to the treatment they receive or the health intervention that is offered is becoming increasingly important. Discussing the risks and benefits of health interventions, treatment options, as well as screening programs, is therefore becoming an essential part of modern-day health care. Adequate communication of the risks implies that information is provided in an optimal way in order to increase understanding of the risks. This is considered to be crucial for appropriate decision-making. Providing people with the right information in the most accessible manner is not easy. The communication of risks is especially difficult [1]. Although the communication of risk information is fundamental to nearly all health interventions, no consensus exists regarding the most effective way to provide people with this information. Moreover, it is not clear which factors in the presentation of information promote effective risk communication.

Research has shown that people in general have great difficulties in understanding risks, especially when expressed as percentages [2]. Unfortunately, this is the most common way to communicate risks in health care. The format in which risks are presented affects people's perception and their subsequent decisions [2], [3]. It has been shown in domains outside health care that risks presented in a frequentistic manner lead to more accurate risk perception and better decisions [1]. Several other formats are proposed to communicate risks, such as verbal scales, risk ladders, population figures or pictographs, scales that relate risks to more familiar risks [4], [5]. There is, however, little – though contradictory – evidence about which format is most effective in communicating risks, and which risk format is preferred by counselees [4], [6].

Gigerenzer et al. [1], [2] showed that people understand frequencies better particularly if risks are expressed as natural frequencies. When using natural frequencies, the reference class is clear, i.e. the relation of the numerator with the denominator is clear. When there is a 5% chance of side effects with medication, it is not always clear that this means that out of 100 people 5 will experience side effects, while this is clear when presenting the risk as 5 out of 100 persons will experience side effects. Although the communication of risks through numbers appears to contribute to a more transparent and neutral communication process, it has been shown that the interpretation of numerical risks is not straightforward since individuals translate numerical risks in different verbal categories [7].

Frequencies can also be presented as figures in, for example, rows of 10 × 10 with, say, 5 black figures and 95 white figures, representing the 5 patients who will experience side effects. This format, known as population figures, pictographs or icons, is increasingly being used to communicate risks. It is assumed that risks presented as population figures are even easier to understand than natural frequencies. As with frequencies, the relation between the numerator and denominator is clear. Moreover, because population figures are a form of visual information, it is assumed that they are processed more easily and might therefore be understood better than numbers [8].

Another aspect of risk information that makes it difficult to understand is its abstract nature. It is argued that because of the abstract nature of quantitative risk information or statistical information, this information will not have a big impact on people's decisions [9]. The more concrete the statistical information is, the more easy it is to imagine; and the bigger the affective impact is on people's decisions, the greater the experienced chance that the event will actually occur [10]. Frequencies are easier to understand and imagine than percentages, and have been shown to have a bigger impact on people's judgments than percentages [9], [3]. As population figures are even more concrete – because frequencies are visualized as human figures or faces – it might be assumed that population figures will have more affective impact than frequencies. If risks presented in different formats are indeed processed differently, this might affect the decisions made.

In an experiment, subjects were presented with health scenarios and risk information in three different formats (percentages, frequencies, and population figures) and asked to evaluate the risks and make a subsequent decision for a close friend. We hypothesized that (1) percentages will be the most difficult to evaluate, and population figures the easiest, (2) population figures will be evaluated as having the highest affective impact (i.e. evaluated as the most frightening, worrisome), and (3) risks presented as population figures will be evaluated as having the greatest chance of occurring. In addition, (4) we hypothesized that the different risk formats will have a different impact on decisions, in the sense that a more affective evaluation of risks would lead to more risk-averse decisions.

Section snippets

Participants

Through advertisements, we recruited 107 students from the Vrije University of Amsterdam. The mean age was 23 and 73% were women. Their major was medicine (16%), health sciences (13%), biomedical sciences (21%), and others. They were given 10 euros for their participation.

Scenarios

Participants were asked to evaluate four health scenarios. These cases were presented in two different sequences to control for order effects. The scenarios were the following (see also Appendix A): (1) Prenatal testing for

Results

There were no significant differences in cognitive evaluation of the risks between formats. Population figures were not evaluated as the easiest format in all scenarios and there were no significant differences between scenarios (see Fig. 1). The affective evaluation of the risks was, however, significantly different between formats (F(2, 104) = 4.16, p < .05; see Fig. 2). Population figures were generally evaluated as having the most impact (i.e. most frightening, worrisome, and serious) as

Discussion

Different risk formats had different effects on respondents’ evaluation of the health risks presented. Contrary to our expectation, population figures were not evaluated as the easiest format. Population figures were shown to have the biggest affective impact and risks presented as population figures were evaluated as significantly greater than the risks presented in other formats. For one of the four decision-making situations the format of the risks presented influenced their decision, and in

Conflicts of interest

There are no conflicts of interest.

Funding: Funding was received from the Dutch Cancer Foundation (C.F. Ockhuysen-Vermey), and had no role in this study.

References (18)

There are more references available in the full text version of this article.

Cited by (0)

View full text