Amit Ghosh, M.D., an associate professor of medicine at Mayo Clinic, says patients are thinking about two things when they’re making decisions about treatment: What is the likelihood of a bad thing happening? And how will it affect my life?

Bookmark and Share

November 2007 | Back to Table of Contents


Decisions, Decisions

Cancer patients are deluged with data about risks, benefits, survival, and recurrence as they weigh treatment options.

Nurse practitioner Lisa Starr knows that a diagnosis of lobular carcinoma in situ (LCIS) presents a woman with a tough set of choices. Should she opt for surveillance, have preventative chemotherapy, or undergo prophylactic bilateral mastectomy to gain peace of mind? All are reasonable options, given that the diagnosis puts a woman at increased risk for developing invasive cancer in either breast later on.

Starr, who works with breast cancer patients at the Duluth Clinic Cancer Center, also knows that to make a good decision, patients need to be fortified with facts—statistics about risk, rates of survival, potential benefits of various treatments, toxicities, and more. But she wonders, If she were sick, how would she want to receive this information?

With that thought in mind, Starr slowly unfolds the statistics in ways she hopes her patients will understand. For example, she tells a patient that an LCIS diagnosis puts her risk for cancer slightly above the lifetime average.

By doing this, Starr is, as she says, “planting seeds” for later discussions that will be loaded with figures. “I try to lay the groundwork ahead of time without so many numbers,” she says.

Proceed with Caution
Nowadays, a cancer diagnosis plunges a patient into a sea of statistics. With a half-
century of data from clinical trials behind us and more coming at us daily, patients as well as physicians must swim through a steady stream of information in order to make decisions about treatment.

Daniel A. Nikcevich, M.D., Ph.D., director of the Duluth Clinic Cancer Center, says patients frequently hand him 20 to 30 pages they’ve printed out from the Internet on cancer treatment options and survival. “Quite commonly,” he says, “what they’ve printed out is completely wrong.” Like Starr, Nikcevich says the key to helping patients sift through the information is to proceed with caution. “If you blow right through it, and you start talking in terms that aren’t recognizable and start saying, Well, the relative risk reduction is this, the numbers needed to treat is this, and the numbers needed to harm is that, then you get into mumbo jumbo.”

Nikcevich’s approach is to systematically walk patients through their diagnosis, the stage of the disease, the available therapies, and whether or not those treatments will provide meaningful benefit.

That may take some time. “Let’s say a woman has a stage 1 breast cancer,” he says, launching into an explanation that shows the breadth of the territory he’ll cover. She can be treated with hormonal therapy that is going to reduce her risk of recurrence by half. So if she had a 20 percent chance of recurrence, she’s gotten the risk down to as little as 10 percent. That 50 percent risk reduction needs to be understood in the context of absolute benefit (the number of patients who must be treated with a therapy in order to gain life expectancy) as well, which is about 10. If the patient opted for chemotherapy, it would add another 20 percent of risk reduction. “That sounds great,” he says. “But then you say, 20 percent of what? Twenty percent of 10 is 2. Then you have to ask, Is it worth getting chemotherapy, losing your hair, risking a potentially life-threatening infection, and having nausea, fatigue, and potential cardiac issues all to gain approximately 2 percent in risk reduction?”

How Much Gets Through?
Research is showing that what people understand about such information varies. “Depending on how much time the physician is spending and what tactics they’re using to break [the news about] the risks and communicate about the treatments and options, patients’ reactions and understanding of the risk factors are very variable,” says Amit Ghosh, M.D., an associate professor of medicine at Mayo Clinic who researches how physicians communicate with patients about risk. Ghosh notes that how patients perceive risk is based on their life experience, their reading, and their friends’ experience. Starr would agree. People are sometimes making decisions based on things in their history that they can’t articulate or even identify, she says.

A patient’s understanding also might be affected by his or her level of literacy. With an estimated 90 million U.S. residents reading at the fifth-grade level or below, many people don’t know basic medical or mathematical concepts. “It’s affecting their ability to understand complexity,” Ghosh says. “And you can’t imagine the complexity of an oncologic consultation and the variety of decisions that need to be made.”

Conversation Starters
University of Minnesota pediatric oncologist Joseph Neglia, M.D., M.P.H., who often sits before distraught parents who’ve just learned their child has been diagnosed with cancer, says the informed consent process requires a depth of understanding that’s hard to attain. For a diagnosis of acute lymphoblastic leukemia, the most common type of cancer in children, he’ll have to explain at least six different therapeutic pathways.

Missing the Point

Daniel A. Nikcevich, M.D., Ph.D., director of the Duluth Clinic Cancer Center, says patients aren’t the only ones to stumble over the numbers emerging from cancer studies. He says physicians often fail to discern whether reports of improvement due to one therapy or another are clinically meaningful.

He cites as an example a paper about the drug panitumumab, a monoclonal antibody that was recently approved by the FDA for treatment of patients with metastatic colon cancer. Researchers reported a statistically significant difference in time to progression between the group taking the drug versus those who received supportive care. “The stats were significant,” Nikcevich says, “except that the difference in absolute terms was .7 weeks.” He pauses, then clarifies that that’s five days. “Is that clinically meaningful?” he asks, also mentioning that the drug costs thousands of dollars a month. “As a physician, I have to say, What are we getting out of that?”—C.P.

Neglia says he does several things to help parents. As he talks, he writes down the key points he’s making. He also encourages parents to involve friends and relatives. “It’s very hard for a mother and father whose child has just been given a life-threatening diagnosis to think objectively and calmly about it.”

He shies away from using specific numbers when talking about the likelihood of a
cure or the chance of recurrence. “Those numbers are population statistics,” he says, and knowing that a particular cancer has a 90 percent cure rate is not all that useful when it comes to treating a single child. “It means, If I have a room full of 100 children, 90 of them will be cured and 10 won’t. The trouble I have is, I can’t tell who those 10 children are,” he says. “An individual is really 100 percent or 0 in that regard.”

Such conversations take time, Neglia says. “For a major event, it’s impossible to give out the detail of information [needed] in a one-hour or 45-minute office visit.”

To help cancer patients visualize what they’re saying about the risks and benefits of a treatment, physicians sometimes turn to decision aids that use graphics to illustrate mathematical concepts. Ghosh says such aids can be valuable. His research is showing that patients prefer frequency formats, which use highlighted human figures to show risk estimate, over bar graphs. Bar graphs, he cautions, can be problematic: A tiny percentage can look like a big gain depending on the reference scale used in the graph. “They can be manipulated to serve the interests of a drug company or whatever,” he adds.

Ghosh says any conversation about risk ought to address the nature of the decision, treatment options, the known risks and benefits of each option, what’s uncertain, what the patient is perceiving, and the patient’s preference. Mayo Clinic doctors are encouraged to use an approach represented by the mnemonic CARE: Cite the data in general terms, Add numbers and probabilities, Reinforce using a visual aid, and Express encouragement and hope.

Surprisingly, Ghosh adds that a physician may be most effective in the moments between those steps. “There’s power in silence,” he says, suggesting that physicians allow patients time to digest what they’ve just heard before continuing the conversation. “Don’t expect an answer in five seconds. There’s a magical thing about quietness that is very important.”—Carmen Peota


. .