Reducing the Risk That Patients Get It Wrong
Article Outline
See article on page 429.
Patients increasingly face a bewildering array of health risks. Deciding which risks are important is hard—but it matters. Inaccurate risk perceptions have important consequences. If people at high risk of disease are unaware of their elevated risk, they may fail to appropriately consider beneficial interventions that might help. If those at low risk of disease have a falsely heightened sense of risk, they may experience undue anxiety and may pursue interventions that offer them more harm than benefit. Understanding the magnitude of a health risk is fundamental to deciding whether the risk is worth worrying about, and, if so, whether to consider taking some action to reduce the risk.
In this issue of Gastroenterology, researchers report a study motivated by the belief that risk perception is important in helping patients with Barrett’s esophagus make good medical decisions.1 They asked patients with this diagnosis to estimate their chance of developing esophageal cancer in the next year and compared their answers with their actual risk—which is generally thought to be about 0.5% per year. Eliciting beliefs about events this rare is difficult—standard methods tend to drive respondents to overestimate (eg, given a 0- to 100-point scale, people rarely write in a fraction less than 1).2 A strength of the study is that the authors used a validated assessment tool that makes it easier to express values less than 1%. Even using this scale, the authors found that two-thirds of these patients overestimated the annual risk of esophageal cancer—with an average perceived risk of nearly 14%, almost 30 times higher than the actual risk. Many of the remaining patients underestimated their risk, highlighting the bottom line: these patients with Barrett’s esophagus do not have an accurate perception of their risk of esophageal cancer.
Why were so many patients wrong? Why did these patients, who carried the diagnosis of Barrett’s esophagus for an average of 5 years, not have a better sense of their esophageal cancer risk? Maybe they were never told.
Why not? Perhaps their physician did not know the 0.5% annual risk figure. But, even when physicians know the numbers, they may not try to convey them to patients: they might believe it takes too much time, or they may worry that the patients just will not understand. It might not occur to them to even try: physicians are not taught to talk to patients about numbers, and studies suggest that quantifying risk is something physicians rarely do.3, 4
To better understand how to talk to patients about risk, physicians need to recognize just how powerfully their use of language can influence what patients hear. In Table 1, we consider how different kinds of statements might affect the perceptions of Barrett’s patients. A simple qualitative statement such as “you are at increased risk of esophageal cancer” is inherently ambiguous—words like increased risk or high risk mean different things to different people. The physician may mean the risk is a little bit higher (and still quite small); the patient may hear that cancer is inevitable. Attaching numbers to risks helps deal with this ambiguity. Unfortunately, the numbers given are often relative risks, such as “Barrett’s patients have a 30-fold increased chance of developing esophageal cancer.” Unless patients are told, “compared with what,” the relative risk statistic can greatly exaggerate the increased risk.5, 6 The increased risk of esophageal cancer with Barrett’s is a perfect example. Here the 30-fold increase translates to the difference between annual risks of 0.02% and 0.5%. Providing the risk for patients with and without Barrett’s gives the complete picture. But 1-year estimates—like the 0.5% statistic—minimize the risk since the chance of esophageal cancer accumulates over time. We think the best approach is to provide 10-year risk estimates: 10 years is a time frame people can easily imagine, it is long enough for risk to accumulate, and also long enough for risk-reducing strategies to have an impact.
Table 1. How Different Ways of Describing Risk Might Affect the Perceived Risk of Esophageal Cancer in Patients With Barrett’s Esophagus
| Category of risk statement | Statement | Effect of statement |
|---|---|---|
| Qualitative | “Patients with Barrett’s esophagus are at increased risk of esophageal cancer” |
Exaggerate risk Patient assumes the risk must be big—why else would the doctor be concerned? |
| Relative risk only | “Patients with Barrett’s esophagus have a 30-fold increased chance of developing esophageal cancer compared with the general population” |
Exaggerate risk This format has been shown to exaggerate how patients5 and physician6 perceive risk—people tend to focus on the big “30-fold increase” and forget to consider the base rate (30-fold greater than what?) |
| 1-year absolute risk | “For every 1000 patients with Barrett’s esophagus, 5 will develop esophageal cancer in 1 year” |
Minimize risk Barrett’s patients are generally followed up for years and the risk accumulates, something the 1-year frame does not address |
| 10-year absolute risk | “For every 1000 patients with Barrett’s esophagus, 50 will develop esophageal cancer in 10 years” |
Neutral The 10-year time frame is easy to imagine, and better captures the idea of accumulating risk; we think this is better than the vague concept of “lifetime risk” (ie, how long is a lifetime?) Some experts suggest adding phrases like the following to help people remember the denominator: “Another way of saying this is that, for every 1000 patients like you, 950 will not develop esophageal cancer over the next 10 years.” |
| 10-year absolute risk for comparison group | “compared with about 2 per 1000 people in the general population” |
Neutral Including the absolute risk in the general population helps people appreciate how much Barrett’s increases risk |
However, even the best statement about the risk of esophageal cancer is not enough; patients also need to know the extent to which the risk can be altered by what they choose to do. In the case of Barrett’s esophagus, patients need to know the chance that they will die from esophageal cancer if they do or do not undergo surveillance endoscopy. Then, to decide whether the benefit of surveillance endoscopy is worth it, patients also need to know about its potential downsides: if the harms are minimal, even a small benefit might be enough; if there are substantial harms, patients may want considerable benefit before they agree to surveillance.
We created an idealized table to present the foregoing information (Table 2). One thing that became obvious is that key data are not readily available. A great deal of basic research will need to be performed to complete the Barrett’s esophagus facts box. However, this is exactly the work needed if we want to help patients with this condition make wise decisions about surveillance endoscopy. There are other things people would want to know, such as whether surveillance affects all-cause mortality or how it impacts on quality of life. But getting the 10-year risk of esophageal cancer death with and without surveillance is a great place to start.
Table 2. Barrett’s Esophagus Facts Box
| What is the chance that a 55-year-old person with Barrett’s esophagus will experience the following in the next 10 years? | ||
|---|---|---|
| Outcomes affected | If he/she does NOT have surveillance | If he/she has surveillance |
| Benefits | ||
 Develop esophageal cancer | 5.0% (50 in 1000) | ? |
| Die from esophageal cancer | ? | ? |
| Harms | ||
| Bleeding in the esophagus after endoscopy | ? | ? |
| Tear in the esophagus after endoscopy | ? | ? |
References
- . The perception of cancer risk in subjects with prevalent Barrett’s esophagus enrolled in an endoscopic surveillance program . Gastroenterology . 2005;129:429–436
- . A new scale for assessing perceptions of chance (a validation study) . Med Decis Making . 2000;20:298–307
- . Informed decision making in outpatient practice (time to get back to basics) . JAMA . 1999;282:2313–2320
- . How do physicians talk with their patients about risks? . J Gen Intern Med . 1994;9:402–404
- . The framing effect of relative and absolute risk . J Gen Intern Med . 1993;8:543–548
- . Absolutely relative (how research results are summarized can affect treatment decisions) . Am J Med . 1992;92:121–124
Supported by a Research Enhancement Award from the Department of Veterans Affairs. S.W. and L.M.S. were supported by Veterans Affairs Advanced Research Career Development Awards in Health Services Research and Development and Robert Wood Johnson Generalist Faculty Scholar Awards, and a National Cancer Institute grant (CA 104721).
The views expressed herein do not necessarily represent the views of the Department of Veterans Affairs or the United States Government.The authors thank H. Gilbert Welch, MD, MPH for his thoughtful review of this editorial.S.W. and L.M.S. contributed equally to this work, and the order of their names is arbitrary.
PII: S0016-5085(05)01197-2
doi:10.1053/j.gastro.2005.06.035
© 2005 American Gastroenterological Association. Published by Elsevier Inc. All rights reserved.
Refers to article:
- The Perception of Cancer Risk in Patients With Prevalent Barrett’s Esophagus Enrolled in an Endoscopic Surveillance Program
