Speaking of cancer: Tips on how to convey risks to patients

Genetic test results can be hard to understand, even for physicians. For a patient with no scientific education, especially one who is also facing fear of cancer, comprehension can be much, much harder.

“We're talking about communicating to people who may have never heard of genetics, providing very complicated information with complicated words and terms,” said Mary B. Daly, PhD, FACP, chair of clinical genetics at Fox Chase Cancer Center in Philadelphia.

As information about the genetic roots of disease moves from bench to bedside, internists may be having more of these conversations. The gene mutations in BRCA1 and BRCA2, which significantly increase risk for breast and some other cancers, are an example of how quickly genetic data can move from research to practice.

“From the time they were discovered in 1995, it was only about a year before we had the ability to order tests. Despite the fact that it's a small percentage of cancer and despite the fact that we don't know a lot about it, there's been a lot of exposure,” said Dr. Daly.

She recently conducted a study about how patients understand and communicate with their families about BRCA test results. Although the results are not published yet, Dr. Daly shared some of her insights on the topic with attendees at the American Association for Cancer Research's Frontiers in Cancer Prevention conference, held in Philadelphia in November 2010.

The study participants were people who had chosen to be tested for the BRCA genes. “It's usually, although not always, a woman who either has breast or ovarian cancer or has a compelling family history of it,” Dr. Daly said.

A few months after the results were given, the researchers followed up with the tested patients (known as probands) and their first-degree relatives to see what kind of communication about the tests had occurred. About 75% of the probands had shared their results with at least one relative, Dr. Daly reported.

Some trends emerged in the research about whom the probands talked to about their results. “They didn't share it with all of their relatives in most cases. In the case of hereditary breast and ovarian cancer, women who receive the results are more likely to share it with female relatives than male relatives,” said Dr. Daly.

That may be because the probands perceive their female relatives to be more at risk if the family carries the genetic mutation, although both BRCA1 and BRCA2 have also been associated with increased risks of prostate, pancreatic and male breast cancers.

Perceptions of risk may also explain why patients were more likely to share their results with their adult children than their siblings or parents. A person carrying the mutation has a 50% chance of having passed it on to a child and so might feel it necessary to warn his or her offspring.

Telling a parent about the test result, on the other hand, might feel more accusatory, and that may be why probands were least likely to tell their parents. “They were trying to spare the parent that sense of guilt” for having passed on the gene, said Dr. Daly. In addition, elderly parents might be less able to understand the test results.

The expected reaction of relatives was a determinant of whether patients shared their results. “If they thought that the relative didn't think genetic testing was a good idea, they were much less likely to share that information, even though both sets of relatives are at equal risk of having the mutation,” said Dr. Daly.

Perhaps the most surprising result of the study was that even if probands made the decision to share their results, the relatives didn't always get the message. “Even though the probands reported sharing the genetic test results, there was 22% [of relatives] who didn't remember or didn't think they heard the information,” said Dr. Daly.

And of those who remembered hearing about the test, only 35% could accurately report the result. Relatives were most likely to misreport the result if the proband had an indeterminate or inconclusive result, such as testing positive for another gene variant of unknown significance. “True positives and true negatives were much more likely to be correctly remembered,” said Dr. Daly.

The study's findings show the need to improve communication about genetic test results, which is likely to happen more often in medical practice, predicted Dr. Daly. BRCA1 and BRCA2 are just the tip of the iceberg of genetic information that people are going to have to cull through as we move forward.”

They may not be ready to navigate the iceberg, but researchers have found some methods to increase the chances that patients will correctly understand their risk for disease, according to another speaker at the conference. James G. Dolan, FACP, associate professor of community and preventive medicine at the University of Rochester in New York, summarized some of the evidence on accurately communicating risk.

“Fortunately, we have some simple rules you can follow,” he said. “The first decision is what outcome you're going to talk about, the second is how you're going to express the chance of that outcome.”

That first decision has a well-established answer: Tell patients about both outcomes. For example, if the patient has a 10% chance of developing cancer, also make it clear that he has a 90% chance of not developing cancer. “It's well known that people perceive risks differently depending on how you present the risk,” said Dr. Dolan. “People are more likely to perceive a higher chance of risk and are more likely to take action if all they see is the chance of the bad outcome happening than if they see both the chances that a bad outcome will and will not occur.”

Presenting the chances of both positive and negative outcomes follows one of the other rules of risk communication: Don't make the patient do math. “You should always try to minimize the need that the recipient has to process the data. So if to fully understand the information, they have to subtract something, you should do that for them,” Dr. Dolan said.

Subtraction implies a numerical explanation of risk, which is usually preferable to using only words, Dr. Dolan advised. Patients' comprehension of risk calculations will vary depending on their numeracy, but there's even more variation possible if a risk is described as low, average or high.

“Words can help put the risk in context, but as the primary source, they're not going to do a very good job at reducing the gap between actual and perceived risk,” he said. “If you tell people that the risk is low, there are a wide range of interpretations.”

There is also a range of ways to present a risk numerically. Both percentage and x per 100 formats work well, but 1 in x formats should be avoided (e.g., it's better to say a 5% risk or 5 in 100 than 1 in 20). Stay away from decimals, too. “It's very hard to conceptualize what a 3.1% risk is,” said Dr. Dolan.

In addition to a verbal explanation, any kind of graphic representation will make it easier to conceptualize risk, because it adds the brain's visual processing power to the task. Whether it's a table, flowchart or graph, the graphic should portray the risk as part of a whole. For example, the National Cancer Institute website shows 100 stick figures with three figures highlighted to portray a 3% risk.

But the illustration doesn't have to be so elaborate. “Frequency diagrams actually have been shown to be very effective and seem to mirror the way people naturally think about risk,” said Dr. Dolan.

Some combination of these visual options and numerical explanations should help physicians accomplish the main objective of talking about risk with patients. “The goal is to communicate risks people have to them in a way that they have a very accurate perception, as accurate as possible, of what they're dealing with,” Dr. Dolan said.