One in 1,000 dies: communicating medical numbers

The Collaborative Research Center “Treatment Expectation” focuses on risks, chances of recovery, and side-effect rates in patient education

Goethe University Frankfurt

FacebookXLinkedInWeChatBlueskyMessageWhatsAppEmail

 

image: 

Tobias Kube 

view more 

Credit: André Zelck

FRANKFURT, MARBURG, ESSEN. Similar – yet not the same: Many studies show that patients often struggle to interpret numerical information in medical contexts, especially probabilities related to recovery and side effects. In a recently published Letter in the prestigious journal JAMA (Journal of the American Medical Association), Professors Tobias Kube (Goethe University Frankfurt) and Winfried Rief (University of Marburg) explain which phrasing can help prevent nocebo effects in communication in outpatient and clinical settings.

 

Is “three out of 100” the same as 3%? Mathematically, yes – psychologically, not necessarily. While the numbers are identical, patients often perceive the two descriptions differently. This is one of the key points highlighted by Prof. Tobias Kube (Clinical Psychology and Psychopathology, Goethe University Frankfurt) and Prof. Winfried Rief, Chair of Clinical Psychology and Psychotherapy at the University of Marburg. Rief also serves as deputy spokesperson for the Collaborative Research Center (CRC) “Treatment Expectation,” which investigates placebo and nocebo effects; Kube is an associate member of the CRC. One of the CRC’s central research questions is how healthcare providers can strengthen placebo effects and avoid nocebo effects through medical communication. Which pitfalls should clinicians avoid when discussing risks, chances of recovery, and side-effect rates during patient consultations?

 

Numbers and Numeracy

Many people struggle with basic mathematical tasks. In a study of 4,637 adults in the United States, only 34 percent were able to identify the largest value in an unordered sequence of numbers. Brian Zikmund-Fisher, Professor of Health Behavior and Health Equity at the University of Michigan, reported this in JAMA in late October 2025. The situation in Germany is likely not fundamentally different. Zikmund-Fisher recommends five clear strategies for presenting numbers in an understandable way and advises against vague verbal descriptors such as “frequent,” “very rare,” or “unlikely.” Without context and comparison, such terms offer little guidance and can amplify fear as well as unintended expectation effects.

However, as Kube notes in the recent JAMA commentary, numbers in medical contexts also carry risks. What the U.S. professor did not address in his JAMA article are so-called framing effects. Kube and Rief therefore show that patients can perceive numerically identical test results and probability statements very differently depending on how they are presented.

 

Positive Percentages

“There is no mathematical difference between saying ‘90 percent of patients recover from the infection’ and ‘10 percent do not recover,’ but the first phrasing emphasizes the high likelihood of a favorable outcome,” explains Prof. Tobias Kube (Goethe University Frankfurt). “That is positive framing.” The first formulation tends to be reassuring, while the second can provoke anxiety. “That’s why we should aim for positive framing in clinical practice – especially when communicating potentially distressing information. In such situations, explanations such as the likelihood that a treatment will work or the probability of severe side effects should be placed in a positive context.”

 

The Framing Effect 

This “framing effect” describes why we evaluate the same information differently depending on how it is phrased. Psychologists Daniel Kahneman and Amos Tversky showed in their early, pioneering work on decision-making that two statements with identical content can trigger very different emotional responses – and therefore different decisions – depending on linguistic framing. The same information can thus be experienced differently in perception, emotional state, and evaluation. This is why wording matters: the brain responds not only to content, but also strongly to emotional tone. Kube goes a step further: “Beyond positive versus negative framing, it also matters whether probabilities are presented as percentages or as frequencies.” For example, the statement “one in a hundred will die” often feels far more alarming than “one percent will die.” “When communicating medical risks using negative framing, it is often better to use percentages, because patients perceive them more abstractly. That makes it less likely they will immediately imagine themselves as the one person out of a hundred who is affected,” advises Prof. Rief (University of Marburg).

 

The Recommendation

Numbers are a central component of medical communication, but they should be chosen carefully with regard to framing effects. Patients who are highly anxious require special attention and expanded communication. “Highly anxious and very concerned patients should be given a detailed explanation of how these numbers are to be understood,” explains Kube.

 

Conclusion 
“Many patients do not fully understand what a doctor has said – or, more importantly, what was meant – during consultations,” confirms neurologist Prof. Ulrike Bingel, Head of University Pain Medicine at Universitätsmedizin Essen and spokesperson for the CRC “Treatment Expectation.” She adds: “Effective health communication primarily requires time – and that time is often lacking in clinical practice.”

Especially given tight time resources in the healthcare system, Kube sees substantial potential in careful risk communication: “Positive framing costs nothing and does not require additional time in consultations – which makes it particularly easy to implement.”

“Every patient asks about risks and benefits because information provides reassurance. That’s precisely why we need to train therapists and clinicians to communicate diagnoses, treatments, and potential side effects in a way that is sensitive to expectations,” emphasizes Bingel. Current studies provide clear guidance on what matters. Patients should not be left to search for explanations online on their own, Bingel warns.

---

Subject of Research

People