When a doctor describes a procedure as “rarely associated with risk”, what does that actually mean? Words like “common,” “unlikely,” or “rare” may sound clear, but experts in medical decision-making warn that leaving out numbers can mislead patients about how big a risk really is.
In a paper published on April 29 in the Journal of General Internal Medicine, a team of researchers and clinicians explained that patients often overestimate risks—such as side effects or disease complications—when they receive only verbal descriptions. The authors argue that clinicians should include numbers more often and offer five evidence-based ways to make statistics easier for patients to understand and use.
“One of the purposes of this paper is to help physicians figure out how to communicate numeric information about risks so that patients can understand and use them to make better choices, take charge of their health and be healthier long term,” said Ellen Peters, a professor at the University of Oregon and director of the Center for Science Communication Research.
Peters draws on about two decades of research into how people interpret numbers and make health decisions. She noted that clinicians often worry patients will not understand numeric information, especially since many adults struggle with basic numeracy. However, her research suggests that many people actually prefer receiving numbers and tend to view messages as more trustworthy—and the speaker as more expert—when data is included.
Working with physicians Paul K.J. Han of the National Cancer Institute and Clara N. Lee of the University of North Carolina, Peters hopes the recommendations will support better shared decision-making between patients and clinicians.
“There’s a whole raft of strategies that you can use, some of which might be more appropriate in one situation versus another,” Peters said. “But by choosing one of them, you can help people use statistics more than they typically would. Otherwise, numbers are just abstract and meaningless.”
Here are the five strategies the authors recommend.
1) Use numbers, not just words
Doctors often rely on verbal labels because many patients find numeric concepts challenging. Still, research indicates that people understand risk more accurately—and respond more appropriately—when numbers are presented alongside verbal descriptions.
Providing numeric estimates can correct a patient’s initial assumptions. When people hear only “common” or “rare,” they may fill in the blanks with their own expectations, which can inflate perceived risk and affect choices.
Do: Headaches are common side effect and occur in 7% of patients.
Don’t: Headaches are common side effect.
2) Make the numbers manageable
When patients feel overwhelmed, they may fall back on shortcuts such as gut feelings, prior beliefs, or stories they have heard from others. To reduce cognitive overload, the authors suggest keeping information focused on what matters most for that patient.
For example, if there are three treatment options but one is clearly not suitable for a person’s specific condition, it may be better not to spend time on the irrelevant option. The paper also encourages clinicians to do the math for patients when possible—for instance, explaining a risk over the total number of years someone expects to use a medication rather than giving only an annual percentage.
Do: Emphasize the key facts and tailor the numbers to the patient’s situation.
Don’t: Include information that isn’t relevant to the decision at hand.
3) Put statistics into context
Numbers without context can be hard to interpret. Peters recommends adding meaning through comparisons or evaluative labels so patients can tell whether a risk is relatively high or low.
Do: 93% of patients survive with treatment A, compared with 99% with treatment B.
Don’t: 93% of patients survive with treatment A.
4) Acknowledge uncertainty
Risk figures are estimates, not guarantees. In some situations, being explicit about uncertainty can help patients understand what a probability does—and does not—tell them. For example, knowing that a risk is 40% does not reveal whether a particular individual will end up in the 40% or the 60%.
Do: These numbers are our best estimate based on current research. We don’t know your exact personal risk because there are factors about you that may not have been studied or fully understood yet.
Don’t: Present a numeric risk as perfectly precise or certain.
5) Check understanding with the teach-back method
Experts often overestimate how clearly they have communicated and how much others understand. The teach-back technique helps by asking patients to explain—in their own words—what they understood as the key points. This gives the clinician a chance to correct misunderstandings and reinforce what matters.
Do: This can be hard to take in, and I want to be sure I explained it clearly. Can you tell me how you understand the pros and cons of taking drug X?
Don’t: What questions do you have?
The authors note that these approaches can also help patients advocate for themselves. If a doctor mentions a possible side effect, patients can ask how often it happens. If they are presented with too many possibilities at once, they can request a simpler summary focused on the most important points.
Peters said she plans to test which of the five approaches is most effective, and she is also interested in how storytelling and anecdotes influence patient decisions.
“Physicians have very limited time in any one appointment and are often faced with a similar patient over and over,” Peters said. “If you’re trying to help them better communicate with patients, you’ve got to provide things that are fast and scripted so that everyone can make informed decisions about their care.”