How to Better Communicate Health Risks and Improve Health Outcomes: Probabilities and Proportions
Part I ─ Probabilities and Proportions
“If we don’t intervene, then you’ll die. If we do intervene, the risk of complications is very low.”
Sometimes decisions about health risks are easy. Administering CPR to save a life makes sense. But sometimes the risks result from behaviors a person is taking or a complicated combination of factors, as with diabetes or a heart attack. How do people think about and respond to these risks?
What if you could reduce the risk of getting diabetes by 50 percent by losing weight? What sort of trade-offs would you be willing to make? What other information would you want before making that decision? How would you feel? Understanding risk matters for decisions about medication, surgery, and our everyday behaviors when it comes to our health.
Understanding risk, probabilities, and percentages can be difficult. This difficulty is compounded when the risks are related to outcomes like death or serious health complications at some point in the future that result from a number of different factors that accumulate over time. What does it mean to halve risk when the baseline rate is low? People may struggle to process risk information the way it is normally presented. Here’s why:
Psychological pitfalls of risk communication
Understanding risk requires interpreting percentages and probability, and this requires “numeracy.” Sixty-four percent of the adult population in the U.S. has limited math skills that don’t include dealing with proportions and probabilities – that is, over 6 in 10 adults would be unable to meaningfully interpret the statistic that started this sentence. But percentages and probabilities are a huge part of how health information is presented. People cannot make informed decisions about their health if they cannot understand the basic numbers associated with a risk.
Relative risk is not the same as absolute risk. People with a BMI over 40 are 18 times more likely to develop type II diabetes than someone with a BMI in the normal range, but that ultimately translates into about 36 percent of obese people having type II diabetes. Even with the elevated risk levels, most obese individuals (two in three) will not have type II diabetes. Is it more important for people to understand relative risks or absolute risks?
Solutions to help people respond better to risk information
Make the numbers simple
Instead of communicating both relative risk and absolute risks, it can be more informative to only talk about basic percentages: “two percent of normal weight men have type II diabetes but 37 percent of obese men have diabetes.” Or, since even percentages are hard to understand, we could say “In a group of 100 normal weight men, 2 will have diabetes. In a group of 100 obese men, 36 will have diabetes.”
Use visual aids
Even after moving to absolute terms, asking people to compare numbers and do a lot of mental math can be tricky. To the extent that you can do the math for people and make it visually compelling, you should.
One of the most effective ways to convey risk and relative risk is via a pictogram. Take the example of weight loss and diabetes risk. Each pound of weight loss results in 7.3 percent reduction in risk. But what does that really mean? For a very obese individual, losing 35 pounds could cut diabetes rates in half, lowering prevalence from 36 individuals in 100 to just 18. A pictograph can demonstrate this without any numbers or computation.
Making the numbers clearer and easy to understand is a great first step in improving risk communication.
Lindsay Juarez – Senior Behavioral Scientist
Lindsay Juarez is a senior behavioral scientist in the Center for Advanced Hindsight at Duke University. She has a Ph.D. in social psychology and specializes in goal pursuit and self-control. She works with the Envolve Center’s Behavioral Economics (BE) team, which incorporates BE and social science into health-related behavioral modification programs.
Julia O’Brien – Principal Behavioral Scientist
Julia O’Brien is a principal behavioral scientist and leads the Better Living and Health Group at the Center for Advanced Hindsight at Duke University. She has a Ph.D. in social psychology and a background in product research. Julia loves behavioral science and believes it has the potential to solve lots of the world’s problems – especially when paired with technology and bold ideas.
 Schwartz LM, Woloshin S, Black WC, Welch HG. (1997). The role of numeracy in understanding the benefit of screening mammography. Ann Int Med., 127, 966–72.