[Hopeful]

OncologySTAT Editorial Team. 2011 Nov 3, Intervew by L Scott Zoeller

Angela Fagerlin, PhD, is Associate Professor of Medicine and Co-Director, Center for Bioethics and Social Sciences in Medicine at the University of Michigan and a research scientist at the VA Ann Arbor Center for Clinical Management Research.

OncologySTAT: Would you talk about some of the challenges that oncologists and hematologists face when communicating treatment options and associated risks with their cancer patients?

Dr. Fagerlin: I think there are a number of communication challenges that oncologists and hematologists face. The first is that they are talking to patients who probably just found out that they have cancer, which means that the patient is experiencing numerous emotions and, as a consequence, his or her ability to cognitively process the information is much lower. Even if you’re talking to the dean of a medical school, who knows everything there is to know about different types of health conditions, that person won’t be able to fully process all of the information you are providing. This is why it is so important for oncologists and hematologists to make the information as simple as possible when talking to patients. The patient will probably not be able to take in everything he or she is being told at the same level that he or she normally would be able to process such complex information.

The best approach includes two different components. One of which is talking in as plain language as possible. For example, try to avoid using jargon. This will help people better understand what is being said. The other is to present numbers as clearly as possible to patients. There are a lot of numbers to talk about when you are explaining to patients the risks and benefits associated with various treatments and screening options. Here, I’m focusing on the emotions involved when the decision to be made concerns treatment, not screening. It is really important to present the numbers in the simplest way possible. My article in JNCI¹ discusses, in detail, a number of ways to simplify those numbers a little bit more for people to understand them. For example, it is better to give absolute risk information rather than relative risk information. Let’s say you have a treatment that can reduce a woman’s risk of having a recurrence of breast cancer from 5% to 2.5%. The patient could be told that she would have a 2.5% risk reduction if she took the treatment, or she can be told that there is a 50% relative risk reduction. Patients don’t often understand what that 50% risk reduction means. They don’t know what their actual risk is right now, so it can be very confusing for them to understand what the actual benefit of the treatment is when information is presented in terms of relative risk. Consistently presenting information in terms of absolute risks would really help patients understand their risk and the benefit the treatment provides them.

OncologySTAT: Based on your research, what are the most effective methods for communicating complex information to patients?

Dr. Fagerlin: Most of our research has focused on how to present the numbers. There have been a lot of really great scholars who have looked at presenting words more simply, but that is not my expertise. So, I’m just going to focus in this conversation about how to present the numbers more simply.

One method is using pictographs, which are also called icon arrays.^2–4 Basically, a pictograph is a 10 x 10 matrix of rectangles, or some people use smiley faces, in which the rectangles are different colors, depending on the treatment and the risks associated with it. The pictograph shows exactly how many people are affected and how many are not affected, so it shows how many people have a risk compared with how many don’t, which can make that number much more concrete for the patient.

Data are mixed about presenting information in terms of frequencies or percentages. You can say either 5% of people have side effect X, or you can say 5 out of 100 people have side effect X. Some studies have shown that using frequencies help people understand the information better, or it changes their risk perceptions in some kind of meaningful way.^5–7 Other studies have found that there are no differences between perceptions of frequencies and percentages.⁸ I always use both when I’m presenting numbers because I think there are some people who can only understand frequencies. It is easier for them to imagine 100 people in a room, and they can imagine five people being affected by side effect X. But, for people with relatively good numeracy skills, using percentages is just fine. So, I often present both, just to make sure that most people can understand the numbers I am presenting.

Another important thing to do when talking with patients is to make clear the difference between baseline risks and risks due to treatment. Let me give you an example. Both tamoxifen and raloxifene can reduce the risk of developing a primary breast cancer by 50%, but each drug has a lot of side effects.

Some of these side effects are health problems that many women often have, for example hormonal symptoms. Many women in their 40s, 50s, and 60s are going through menopause. They have hot flashes and they have vaginal dryness. If you simply tell the patient that—and I’m making up this number—the likelihood of having hot flashes and vaginal dryness is 85%, she will think that the drug is responsible for all 85% of that likelihood, which is not true, because maybe the patient already had a 65% chance of having hot flashes and vaginal dryness.

It is really important to say that, with this particular drug in this particular example, the likelihood of experiencing hormonal symptoms as a side effect of the drug is 20%. We show this in the JNCI article really clearly,¹ graphically, using different colors in the pictograph to make people understand the risk due to the treatment vs the total risk a woman could have based on her baseline risk.

I think many physicians, actually many people in general, are under the assumption that people should not be given numbers because people can’t understand numbers. And these physicians want to rely only on verbal expressions of numbers, such as you have a low risk, a medium risk, or a high risk. However, that approach is really unhelpful because what a low risk is to you might be a high risk to me.

In addition, what is an acceptably low risk in one context, even to a single individual, might be unacceptable in another. For example, if I’m told that I have a 1% chance of having a sore arm after a flu vaccine, I consider that to be a very low risk, and it doesn’t faze me. If, however, I am pregnant and I am told that I have a 1% chance of having a miscarriage following a CVS, that 1% seems like a significant risk. My point is that using only verbal expressions to convey data can be really confusing and not actually provide good information to patients.

When you have numbers to present, try to use them as much as possible, but present them in a way so people understand what they mean and why they are important. This is an area in which the teach-back method might be really helpful— the physician explains a topic and then asks the patient to teach that topic back to him. For example, the physician asks the patient, “When I say such and such, what do you understand? And what do you think your risks are?” And, when the patient responds, the physician makes sure that the patient, does, in fact, understand.

One differentiation point is that for any given patient, the outcome turns out to be either 0% or 100%, whether he or she has the side effect or not; the patient either survives or doesn’t. It is really hard to reconcile that with the fact that, when people are making decisions, they don’t know if they’re going to have the 0% or the 100% outcome. So, you have to try to give them the best quantitative information you can during the decision-making process. This has to do with the area of psychology that Valerie Reyna, from Cornell University, has written about, which is the idea of “gist vs verbatim”—knowledge and understanding in decision-making.⁹

We try to present everything in a verbatim way, with exact numbers, so that people get the information, but a lot of people then code that information using gist representation. They may not remember 7%, but they have in mind that it’s a low risk. Patients often will change the numbers into the verbal expression, into words, and they know what those words mean to themselves. They know what they personally consider to be low and high risk. The physician will not know what the patient considers to be a low risk, but the patient herself knows what she considers to be a low risk (eg, whether or not she will feel pain), so it’s a very tricky thing.

There are some cases in which numbers aren’t always useful, but I think when people are trying to make decisions about what to do, weighing the risks and benefits, the numbers are helpful. One of the things that I’m really interested in studying and learning more about is how many numbers are too many numbers. I conducted a study recently where knowledge was quite a bit poorer than I would have liked to have seen, and I think it’s because we presented too many risks and benefits, and patients just couldn’t comprehend all the information they received. We don’t really know how much people can take in one sitting, even if the information is in a booklet, which they can go back to. There might be a point where there is just too much cognitive overload, and they just can’t process information. That’s something we don’t know yet, but, although we do have to be thoughtful about when and how we present numbers, presenting numbers is beneficial.

OncologySTAT: How does the way in which a physician presents information influence the patient’s decision making, either in a positive or a negative way?

Dr. Fagerlin: One example of that is presenting relative risk information vs absolute risk information. Research has shown that for both patients and physicians, when relative risk is presented, people are more likely to want treatment, or, in the physicians’ case, more likely to recommend treatment. In two studies,^10,11 investigators conducted an experiment in which physicians were given a chemotherapy scenario that used relative risk information. The results showed that when the benefits of chemotherapy were presented in terms of relative rather than absolute risks, more physicians said they would recommend to a patient that he or she should have the chemotherapy. Similarly, when patients received information in terms of relative risk, they believed the treatment to be more effective or they were more likely to want to start the treatment.

I think that how information is presented can really bias people’s decisions. In these cases, people are not necessarily making a better decision; the presentation of information is affecting their decisions.

OncologySTAT: Are there strategies for helping patients retain more of the information they receive?

Dr. Fagerlin: That’s another area that I’m really interested in investigating some more, although right now I don't know of any current good data on this question. I think there are some obvious things, for example it is really helpful for patients to tape-record the appointment if their physicians allow it. This might be especially true when the patient has just received a serious diagnosis, for example, a cancer diagnosis. Often, in these situations, while the physician is talking about treatment options, patients are not able to process their doctor’s words as they are thinking, “I have cancer. Am I going to die? What’s going to happen to my children? Am I going to lose my hair? What’s going to happen to me?” A patient is thinking all of those things while the physician is talking, and the patient does not comprehend much of the information. Tape recorders are often very useful, but some physicians may not welcome them in the clinic visit. However, in some big cancer centers, patients can actually get access to their records, which summarize all of the visits.

It can be very helpful to have somebody at the appointment with you. That person will think of questions that you might not be considering and help take notes. Taking notes can also be very helpful. These are things that could potentially help the patient retain more of the information.

OncologySTAT: How are you relating this to what you already know and what’s being said now?

Dr. Fagerlin: That’s a really important point. In cancer, a lot of people don’t have any background knowledge. They’ve never had cancer before, hopefully, and so all this information is new. If you don’t know anything about lung cancer, or liver cancer, or kidney cancer, and your doctor is giving you a lot of new information, but you have no previous knowledge to relate it to, it will be harder for your brain to process it and to remember it. In psychology, we say that you don’t have a schema for this, so it’s all new information. It’s really hard to retain all new information because you don’t have anything to relate it to or to help you remember it. If you find out that you have breast cancer, even if you have had experience with people who have breast cancer, it’s new as your own health condition, and everybody’s disease and treatment plans are different.

So even if you’ve been through the journey with someone else, and you might be able to remember some of the information better because you’ve heard the stories and you’ve talked to people about it, it’s still all new information for you. You know what your stage is, how big your tumor is, whether it’s ER-positive, ER-negative—that’s all new information.

OncologySTAT: What other information would be helpful?

Dr. Fagerlin: You know, there’s a great statistic by Isaac Lipkus and his colleagues,¹² who wrote a paper published in Medical Decision Making in 2001. They interviewed hundreds of college-educated adults and gave them a numeracy test. All of the participants in the study had, I believe, at least, a bachelor’s degree, or greater, and about 20% of them didn’t know what was a bigger risk: 1 in 10, 1 in 100, or 1 in 1000.

Similarly, about the same percentage of participants did not know what was a bigger risk of something bad happening: 1%, 5%, or 10%. So, when you think all you have to do is give information and tell somebody, “Well, your risk is 7%,” and they’ll understand and you’ve done your due diligence, it’s not true. These were college-educated adults, so could you imagine what the results would have been with high-school–educated adults or people who didn’t finish high school?

People often just need a little extra help. It’s not enough just to throw out a bunch of numbers. You have to make sure that people understand them, can interpret them, and can use them in their decision making.

References
1. Angela Fagerlin, Zikmund-Fisher BJ, Ubel PA. Helping patients decide: ten steps to better risk communication. J Natl Cancer Inst. 2011;103:1-8.
2. Tait A, Voepel-Lewis T, Zikmund-Fisher B, Fagerlin A. The effect of format on parents’ understanding of the risks and benefits of clinical research: a comparison between texts, tables, and graphics. J Health Commun. 2010;15(5):487-501.
3. Hawley ST, Zikmund-Fisher B, Ubel P, et al. The impact of the format of graphical presentation on health-related knowledge and treatment choices. Patient Educ Couns. 2008;73(3):448-455.
4. Fagerlin A, Wang C, Ubel P. Reducing the influence of anecdotal reasoning on people’s health care decisions: is a picture worth a thousand statistics? Med Decis Making. 2005;25(4):398-405.
5. Hoffrage U, Gigerenzer G. Using natural frequencies to improve diagnostic inferences. Acad Med. 1998;73(5):538-540.
6. Schapira MM, Nattinger AB, McHorney CA. Frequency or probability? A qualitative study of risk communication formats used in health care. Med Decis Making. 2001;21(6):459-467.
7. Peters E, Hart PS, Fraenkel L. Informing patients: the influence of numeracy, framing, and format of side-effect information on risk perceptions. Med Decis Making. 2011;31(3):432-436.
8. Cuite C, Weinstein N, Emmons K, Colditz G. A test of numeric formats for communicating risk probabilities. Med Decis Making. 2008;28(3):377-384.
9. Reyna VF. A theory of medical decision making and health: fuzzy trace theory. Med Decis Making. 2008;28(6):850-865.
10. Chao C, Studts JL, Abell T, et al. Adjuvant chemotherapy for breast cancer: how presentation of recurrence risk influences decision-making. J Clin Oncol. 2003;21(23):4299-4305.
11. Bobbio M, Demichelis B, Giustetto G. Completeness of reporting trial results: effect on physicians’ willingness to prescribe. Lancet. 1994;343(8907):1209-1211.
12. Lipkus IM, Samsa G, Rimer BK. General performance on a numeracy scale among highly educated samples. Med Decis Making. February 2001;21(1):37-44.


Hopeful