© 1999 by Oxford University Press
Journal of the National Cancer Institute Monographs, No. 25, 173-176,
1999
© 1999 Oxford University Press
Challenges to Improving Health Risk Communication in the 21st Century: a Discussion
Affiliation of authors: Department of Psychology, University of Waterloo, Ontario, Canada.
Correspondence to: Geoffrey T. Fong, Ph.D., Department of Psychology, University of Waterloo, Waterloo, ON N2L 3G1 Canada (e-mail: gfong{at}watarts.uwaterloo.ca).
 |
INTRODUCTION |
|---|
 Top Introduction How Should We (or...What Format Is Most...How and Why Does...Can Researchers and Journalists...How Can We Best...NotesReferences |
|---|
Risk communication has always been fraught with challenges, and there is perhaps no greater example of the range and complexity of the challenges involved in risk communication than in the domain of health risk communication and, more specifically, in the domain of cancer risk communication. In this paper, we discuss some important issues that we believe should be addressed to meet the many challenges that face researchers, health communicators, and the media in communicating health risks to patients, to their families, and to the public.
|
HOW SHOULD WE (OR SHOULD WE) COMMUNICATE THE UNCERTAINTY ASSOCIATED WITH RISK INFORMATION? |
|---|
|
TopIntroductionHow Should We (or...What Format Is Most...How and Why Does...Can Researchers and Journalists...How Can We Best...NotesReferences |
|---|
Uncertainty is the bane of a risk communicator's existence, and this is particularly true for those who communicate risk in a domain in which the knowledge base is changing rapidly, such as cancer risk: Not only is the risk estimate expressed in terms of uncertainty, such as a probability, but also the risk estimate itself is subject to uncertainty. Consider, for example, the wide range of estimates for lifetime risk of ovarian cancer among women with BRCA1 mutations. The estimate from one study of BRCA1-linked families (1) was 30%, but with a 95% confidence interval of 8%-47%; the estimate from a population-based study (2) was 16%, with a 95% confidence interval of 6%-28%. A commonly cited estimate of lifetime risk is 50% [e.g., (3)], and a recent text (4) summarizes the literature by suggesting that the lifetime risk is somewhere between 10% and 80%! Given the inherent challenges of generating valid population risk estimates from either cancer registry families or special populations (e.g., Ashkenazi Jews), along with the instability in estimates derived from small samples, it is not surprising that there is considerable uncertainty in risk estimates for many hereditary cancers.
But how then should the risk communicator convey this risk information to the individual (or to the public)? Does one present the estimate from the single, presumably most relevant or highest quality study, if that is even ascertainable, or does one present an average of estimates? As Schwartz et al. (5) point out, communicating a single risk estimate without uncertainty implies a greater degree of precision than is warranted, and they advocate using confidence intervals to capture this second-order uncertainty, also known as "ambiguous" or "vague" probabilities (6). However, presenting a range of risk estimates, rather than a single risk estimate, creates two problems. The technical problem is that it is unclear which range should be presented. Should we present the range of mean estimates across studies (e.g., "16%-30%") or the range of confidence intervals across studies (e.g., "6%-47%")? The communication problem is that presenting a range of risk estimates may evoke negative reactions of confusion and lower trust (7), leading recipients to feel that they have not received sufficient information to assist them in their decision process (8). Telling a woman that her risk of ovarian cancer is between "10% and 80%" seems to say, "Our state of knowledge is so poor that we really don't know what your risk is."
We believe that this is the basic dilemma facing the health risk communicator—one that has both ethical and pragmatic aspects: a conflict between our desire to accurately convey the limits on our present knowledge, which pushes us in the direction of presenting a range of risk estimates, and our desire to provide risk information that empowers recipients to make confident risk-relevant decisions, which pushes us in the direction of presenting a single "best" estimate. How this conflict is resolved will have a powerful effect on an individual's or the public's reactions and will have implications for every aspect of the risk communication itself. For example, the complexity associated with multiple estimates might well make graphical presentations unwieldy and confusing (imagine representing confidence intervals with a series of pie charts). There is a need for research on the effects of vague probabilities on the judgments and decisions about risk and on recipients' trust and confidence in the risk information and in the risk communicator. Such research could help guide those critical initial decisions about how to present risk information to patients and to the public.
|
WHAT FORMAT IS MOST EFFECTIVE IN COMMUNICATING RISK INFORMATION? |
|---|
|
TopIntroductionHow Should We (or...What Format Is Most...How and Why Does...Can Researchers and Journalists...How Can We Best...NotesReferences |
|---|
Lipkus and Hollands (9) provide an excellent review of the existing literature on the effects of visual or graphical displays of risk information. Although research on the cognitive processing of graphs [e.g., (10-12)] has been applied to understanding comprehension and responses to graphical presentations of risk [e.g., (13,14)], we still know little about what graphical methods could lead to more effective health risk communications. In fact, we know little about how (or even whether) graphical presentations are superior to other formats, such as numeric or verbal formats. Because graphical presentations can be so powerful, it is particularly important for research to discern the best methods for presenting information—methods and formats that will maximize comprehension of the message. In short, if a picture is worth a thousand words, we need to ensure that it is conveying the appropriate words.
|
HOW AND WHY DOES TAILORING WORK? |
|---|
|
TopIntroductionHow Should We (or...What Format Is Most...How and Why Does...Can Researchers and Journalists...How Can We Best...NotesReferences |
|---|
Recent advances in new interactive technologies now make it possible to create powerful and sophisticated methods for risk assessment, for risk communication, and for delivering interventions to reduce health risk. The review by Strecher et al. (15) illustrates so clearly that these technologies represent the engine that will revolutionize the nature of risk communication and health interventions, in research as well as in practice.
These new interactive technologies allow the construction of tailored interventions that are much more fine grained and elaborate than has been possible in the past. The review of the literature by Rimer and Glassman (16) provides fairly good evidence for the effectiveness of tailoring, although many of the studies cited do not demonstrate the beneficial effects of tailoring per se because they confound tailoring with other aspects of the intervention (e.g., quantity and specificity of the information). Rimer and Glassman also provide a nice classification scheme for thinking about the dimensions on which tailoring could be accomplished and offer an excellent set of suggestions for future research on tailored risk communications. We wish to amplify one of their points and add another. First, it is important for research to identify which of the many variables that could be employed in a tailored intervention will actually lead to increased effectiveness. Interactive technologies have given us methods for tailoring on virtually any dimension, and now we must engage in the hard work of finding out which variables are worth tailoring on and whether the benefits outweigh the costs. If this interactive technology is the engine, well-designed research should be the navigator that will guide us toward the creation of more effective interventions.
The second point, which follows from the first, is that it is important for us to identify those theoretical models that are best suited for guiding the creation of tailored interventions. One of the attractive features of stage theories, such as the transtheoretical model (17,18) and precaution adoption theory (19), is the assumption—sometimes explicit and sometimes implicit—that interventions that are tailored to an individual's stage of change will lead to greater behavior change. The proper research design for testing this hypothesis is to compare the effects of a stage-appropriate intervention with those of a stage-inappropriate intervention. For example, to properly test the matching hypothesis in the transtheoretical model, an intervention that is tailored for contemplators should be given not only to contemplators but also to individuals at other stages (e.g., preparers). Such fully crossed factorial designs (i.e., those that cross stage of intervention with stage of participant) have not typically been employed [for a notable exception, see (19)]; thus, there still remains a surprising paucity of empirical evidence for the superiority of matching intervention to behavioral change stage, despite the strong intuitive appeal of that hypothesis.
Rimer and Glassman (16) urge researchers to employ factorial designs to identify components of a risk communication that may be most effective. In addition, we urge researchers to measure the theoretically relevant mediators of behavior change so that we can understand how and why a risk communication is effective. In the transtheoretical model, for example, the processes of change are theorized to be the psychological substrate from which the individual progresses from one stage of change to the next. If so, the effects of a matched intervention on behavior change (or progression from one stage to another) should be accompanied by effects on the processes of change associated with the current stage and the next stage. Measuring these mediators within a longitudinal design would allow researchers to perform a rigorous empirical test not only of the intervention but also of the theoretical assumptions of the model (20,21). In a recent example of this methodological strategy, Wang et al. (22) examined the effects of a genetic risk assessment among 60 self-referred patients at an ovarian cancer clinic in a three-wave longitudinal study in which the outcome variable was intentions to engage in future screening behaviors at 2-day follow-up. They found that risk of ovarian cancer that was communicated to patients during the genetic counseling session was significantly related to future screening intentions, and mediational analyses suggested that this effect was because of changes in postsession perceived risk and not because of changes in postsession cancer worry. These results suggest that perceived risk may play a central role in the pathway from risk communication to risk-relevant behavior.
|
CAN RESEARCHERS AND JOURNALISTS WORK TOGETHER TO IMPROVE HEALTH RISK COMMUNICATION TO THE PUBLIC? |
|---|
|
TopIntroductionHow Should We (or...What Format Is Most...How and Why Does...Can Researchers and Journalists...How Can We Best...NotesReferences |
|---|
There has been a tremendous increase in health reporting in recent years. Stories about health are being given more space in the print media and more time in the electronic media. The popularity of health reporting has led to an amplification of both the negative and the positive aspects of health reporting. On the negative side is that some health reporting has followed the same seemingly inexorable descent into sensationalism that has characterized reporting in many other domains. On the positive side is the trend toward more in-depth reporting on important health risks, at least among the "elite media" (23). Both Brody (24) and Levine (25) point out that the public has considerable misapprehension about health risk information, although they focus on different aspects of the problem. Brody provides compelling testimony to the dramatic changes in health reporting and in the public's awareness of health risks that have occurred recently. Her concern that the public is suffering from information overload is well-taken; unfortunately, the overload problem will only become greater in the future, and it is unclear how or whether it can be ameliorated.
Levine (25) focuses on the professional, cultural, and epistemic differences between researchers and journalists and suggests that these differences have created a "chasm" between the two. Indeed, there are a number of differences between the two professions that may conflict. First, differences exist in time frame: Journalists have considerably shorter deadlines and very limited space to tell their story. They have neither the time nor the space to describe the caveats and the nuances of the research. Second, differences exist in epistemic goals and standards, which are nicely summarized by a statement that a journalism professor made to the first author years ago: "Reporters want simple definite statements. They can't stand two-handed researchers—those who say, `on the one hand this . . . on the other hand that. . . .' " But as we know, the researcher's métier is, in fact, "two-handed thinking," e.g., exploring the boundary conditions for a phenomenon, identifying the conditions under which its opposite may obtain (26), or considering alternative explanations. This conflict between the journalist's need to communicate certainty and the researcher's need to acknowledge the uncertainty that is a part of any research endeavor is analogous to the dilemma, described earlier, that is faced by the risk communicator in deciding whether or not to acknowledge the uncertainty of a risk estimate.
Can researchers and journalists work together more effectively to improve media coverage of health risks? We believe that this is possible; however, given the more severe structural constraints imposed on journalists, we suspect that it is the researchers who need to change to a greater extent. Researchers can increase their awareness of the broader implications of their research; they can work together with their public relations office so that press releases describing their findings can be crafted in ways that fairly negotiate between the opposing goals of precision and simplicity. Professional organizations and research funding agencies can offer assistance in the form of workshops and guidelines for communicating with the media.
|
HOW CAN WE BEST MEET THE FUTURE CHALLENGES FACING HEALTH RISK COMMUNICATION? |
|---|
|
TopIntroductionHow Should We (or...What Format Is Most...How and Why Does...Can Researchers and Journalists...How Can We Best...NotesReferences |
|---|
How will the public obtain information about their health risk in the future? Rapid advances in molecular biology, for example, biochip technology (27), will make it considerably easier and less expensive for individuals to obtain information about their genetic risk. As with most technological advances, this will lead to an amplification of both the benefits and the costs. For example, it is doubtful that the safeguards for communicating genetic risk information currently embodied in informed consent and other guidelines will be honored or protected when the risk communication emanates from sources outside the medical research community; thus, the complex legal and ethical issues surrounding genetic testing will only become more problematic.
We are currently undergoing a revolution in how information is disseminated or transmitted to the public by the media (indeed, the very concept of mass media is being transformed). This revolution will lead to dramatic changes in how the public obtains information about health risks. Even today, the vast majority of the public obtains its information about health risks not from the elite media, such as the New York Times or The Washington Post, but from popular television programs or from other non-elite print and electronic sources. What will be the effect of this shift from more controlled, vertical, and hierarchical communication transmission channels to more uncontrolled, horizontal, and nonhierarchical ones? Studies of comprehension of information from physicians suggest that, even under the best of circumstances, a significant proportion of information that is conveyed to patients is either not remembered, misremembered, or misinterpreted. If this is the case, health information that is obtained from the media may be subject to even greater distortion. And looking to the future, with the exponential growth of the Internet, there is a huge and growing number of health-related web pages whose information is of uncertain or dubious validity (28,29). Efforts to create standards for interactive health information (30) are much needed and well intentioned, but the ease with which anyone can create a vivid and compelling web site where health risk information is disseminated without any regard to its validity suggests to us that such standards may ultimately have less impact than they should on ensuring the validity of health risk information on the Internet.
What, then, can be done to counteract these forces that threaten the validity of health-risk information? We have three broad suggestions. First, given the impossibility (and possible unconstitutionality) of regulating content on the Internet and other media, it is necessary instead to fight fire with fire. This can take several forms. For example, institutions such as the National Institutes of Health, the Centers for Disease Control and Prevention, and other governmental entities must endeavor to enhance their reputations as trusted and reliable sources of information about health risk by being proactive. They should monitor the public's concerns about any given health risk, then quickly inform the public about the validity of that health risk, and offer guidelines or suggestions based on the best available sources about the seriousness of the risk and about the methods of reducing the risk, if the risk is serious. In the same way that some corporations have sizable public relations staff that monitor the media for relevant news and provide rapid response for any media report that is worthy of response, so too could agencies provide rapid response for media reports about health risks that are either worthy of amplification or clarification or in need of rejection. Recently, there have been vigorous and concerted efforts by institutions such as the National Cancer Institute to become a more proactive source for health risk information; this is an admirable and important development.
Second, it would be beneficial for guidelines to be created for improving the public's understanding of health risk. Recently, an advisory group convened by the Harvard School of Public Health and the International Food Information Council Foundation issued such guidelines for communicating scientific research on nutrition, food safety, and health. That statement included communication guidelines for scientists; journal editors; journalists; and industry, consumer, and other interest groups, as well as some general guidelines for all parties (31). Creating a similar set of guidelines in the domain of communication of cancer risk, or health risks more generally, might well be beneficial for improving both the communication and comprehension of such risks.
Third, it is important to educate the public about the basics of probability theory and statistics—the language of risk. We concur with the recommendations by Schwartz et al. (5) to develop methods for educating patients so that they have a better understanding of quantitative concepts, such as probability. H. G. Wells's belief that "statistical thinking will one day be as necessary for efficient citizenship as the ability to read and write" has never been so relevant as it is today and will be in the future. Enhancing the public's understanding of probability and statistics would involve additions to school curricula, but other sources could contribute as well. Recent popular books on innumeracy (32) have heightened the public's awareness of the importance of understanding quantitative concepts, and other such efforts are urgently needed.
In summary, we are confident that research will lead to methods for communicating health risks that maximize comprehension and facilitate informed health judgments, decisions, and risk-consistent behavior within controlled settings, such as in genetic counseling or health promotion programs. However, we are less sanguine about whether the same advances can be expected in the uncontrolled environment of the mass media, including the Internet. We fear that much information about health risk that is encountered by the public may be untrue, partially true, or misinterpreted. Incorrect or misleading information may even be deliberately generated by advocates who have an ulterior motive for their actions (28,29). It will take considerable efforts to counteract various forces that threaten the integrity of health risk information.
The stakes are high. We hope that the work of researchers, health risk communicators, and journalists alike will increase our understanding of the factors that contribute to effective risk communication. In so doing, we will be taking important steps toward increasing behaviors and formulating policies that may reduce our health risk in the 21st century.
|
NOTES |
|---|
Supported by grants from the Ontario Cancer Genetics Network and the Social Science and Humanities Research Council of Canada to Dr. Fong. L. A. Rempel and P. Hall were supported by graduate fellowships from the Social Science and Humanities Research Council of Canada.
We thank Cathleen McDonald for her comments on a preliminary draft and Alexandra Fong for her assistance.
|
REFERENCES |
|---|
|
TopIntroductionHow Should We (or...What Format Is Most...How and Why Does...Can Researchers and Journalists...How Can We Best...NotesReferences |
|---|
1 Easton D, Ford D, Bishop DT. Breast and ovarian cancer incidence in BRCA1-mutation carriers. The Breast Cancer Linkage Consortium. Am J Hum Genet 1995;56:265-71.[ISI][Medline]cancerlit;95126121
2
Struewing J, Hartge P, Wacholder S, Baker SM, Berlin M,
McAdams M, et al. The risk of cancer associated with specific mutations of BRCA1 and BRCA2
among Ashkenazi Jews. N Engl J Med 1997;336:1401-8.
3 Mange E, Mange AP. Basic human genetics. Sunderland (MA): Sinauer Associates, Inc.; 1998.
4 Offit K. Clinical cancer genetics: risk counseling and management. New York (NY): Wiley; 1998.
5 Schwartz LM, Woloshin S, Welch HG. Risk communication in clinical practice: putting cancer in context. Monogr Natl Cancer Inst 1999;25:124-33.
6 Budescu D, Weinberg S, Wallsten TS. Decisions based on numerically and verbally expressed uncertainties. J Exp Psychol Hum Percept Perform 1988;14:281-94.
7 Johnson B, Slovic P. Presenting uncertainty in health risk assessment: initial studies of its effects on risk perception and trust. Risk Anal 1995;15:485-94.[CrossRef][ISI][Medline]
8 Fox C, Tversky A. Ambiguity aversion and comparative ignorance. Q J Econ 1995;110:585-603.[CrossRef]
9 Lipkus IM, Hollands JG. The visual communication of risk. Monogr Natl Cancer Inst 1999;25:149-63.
10 Cleveland W, McGill R. Graphical perception: theory, experimentation, and application to the development of graphic methods. J Am Stat Assoc 1984;70:531-54.[CrossRef]
11 Hollands J, Spence I. Judging proportion with graphs: the summation model. Appl Cogn Psychol 1998;12:173-90.[CrossRef]
12 Simkin D, Hastie R. An information processing analysis of graph perception. J Am Stat Assoc 1987;82:454-65.[CrossRef]
13 Stone E, Yates JF, Parker AM. Risk communication: absolute versus relative expressions of low-probability risks. Org Behav Hum Decis Process1994 ;60:387-408.
14 Stone E, Yates JF, Parker AM. Effect of numerical and graphical displays on professed risk-taking behavior. J Exp Psychol Appl 1997;3:243-56.
15 Strecher VJ, Greenwood T, Wang C. Interactive multimedia and risk communication. Monogr Natl Cancer Inst 1999;25:134-9.
16 Rimer BK, Glassman B. Is there a use for tailored print communications in cancer risk communication? Monogr Natl Cancer Inst 1999;25:140-8.
17 Prochaska J, DiClemente CC. Stages and processes of self-change in smoking: toward an integrative model of change. J Consult Clin Psychol 1983;51:390-5.[CrossRef][ISI][Medline]
18 Prochaska J, DiClemente CC, Velicer WF, Rossi JS. Standardized, individualized, interactive, and personalized self-help programs for stages of smoking cessation. Health Psychol 1993;12:399-405.[CrossRef][ISI][Medline]
19 Weinstein N, Lyon JE, Sandman PM, Cuite CL. Experimental evidence for stages of health behavior change: the precaution adoption process model applied to home radon testing. Health Psychol 1998;17:445-53.[CrossRef][ISI][Medline]
20 Fong GT, Jemmott JB III, Jemmott LS. The theory of planned behavior as a mediational model for an HIV risk-reduction intervention for African American adolescents. Proceedings of the Society of Behavioral Medicine's 20th Annual Meeting; 1999 Mar 2-6; San Diego (CA). Ann Behav Med 1999; 21(Suppl): S200-1.
21 West S, Aiken LS. Toward understanding individual effects in multicomponent prevention programs: design and analysis strategies. In: Bryant K, Windle M, West SG, editors. The science of prevention: methodological advances from alcohol and substance abuse research. Washington (DC): American Psychological Association; 1997. p. 167-209.
22 Wang C, Fong GT, Ritvo PG. The psychological effects of genetic risk communication: a longitudinal study. Proceedings of the American Psychological Association; 1998 Aug; San Francisco (CA).
23 Klaidman S. How well the media report health risk. Daedalus 1990;119:119-32.
24 Brody JE. Communicating cancer risk in print journalism. Monogr Natl Cancer Inst 1999;25:170-2.
25 Levine J. Risky business—communicating scientific findings to the public. Monogr Natl Cancer Inst 1999;25:163-6.
26 Salovey P, Schneider TR, Apanovitch AM. Persuasian for the purpose of cancer risk reduction: a discussion. Monogr Natl Cancer Inst 1999;25:119-22.
27
Chee M, Yang R, Hubbell E, Bemo A, Huang XC, Stern D, et
al. Accessing genetic information with high-density DNA arrays. Science 1996;274:610-4.
28
Bower H. Internet sees growth of unverified health claims. BMJ 1996;313:381.
29
Keoun B. Cancer patients find quackery on the Web. J
Natl Cancer Inst 1996;88:1263-5.
30
Robinson T, Patrick K, Eng TR, Gustafson D, and the Science
Panel on Interactive Communication and Health. An evidence-based approach to interactive
health communication. JAMA 1998;280:1264-9.
31
Improving public understanding: guidelines for communicating
emerging science on nutrition, food safety, and health. J Natl Cancer Inst 1998;90:194-9.
32 Paulos J. Innumeracy: mathematical illiteracy and its consequences. New York (NY): Vintage Books; 1990.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||