More or Less Behind the Statistics
More or Less Behind the Statistics

Hannah Fry: Understanding the numbers of cancer

British mathematics professor and broadcaster Hannah Fry has spent many years trying to explain the world through numbers. But when she was diagnosed with cervical cancer she embarked on a new mission – to discover whether the medical world, and we as individuals, make the right choices around treat

Featured Speakers

BBC HostHannah Fry Guest

Topics Discussed

Episode Summary

Executive Summary: The episode explores how statistics can improve life-changing medical decisions, using mathematician Hannah Fry’s cancer experience to show the gap between numbers and patient understanding. It argues that informed consent requires expert translation of risk, not just data, and highlights overtreatment and overdiagnosis through examples from thyroid, breast, and population studies.

Main Topics: Hannah Fry’s cancer diagnosis and treatment decisions (Priority: 5/5): Hannah describes her 2021 cervical cancer diagnosis, radical surgery, and later reflection that she may not have been fully presented with the spectrum of treatment options and trade-offs. Numbers, consent, and patient agency (Priority: 5/5): The discussion centers on how risk statistics should be communicated so patients can make genuinely informed choices that reflect their values and quality of life. A breast cancer appointment illustrating statistical trade-offs (Priority: 4/5): A breast cancer patient is shown being told chemotherapy raises 10-year survival from 84% to 88%, prompting debate over how to interpret marginal benefits in crisis situations. Overdiagnosis and overtreatment in cancer screening (Priority: 5/5): The conversation uses thyroid cancer screening in South Korea to show that finding more cancers does not necessarily reduce deaths, suggesting possible unnecessary treatment. Evidence of hidden or slow-growing cancers in the population (Priority: 4/5): A Danish study and later replications indicate many people carry undetected cancers or precancerous changes that may never become clinically important. The role of doctors as interpreters, not just statisticians (Priority: 4/5): The episode emphasizes that clinicians should translate numbers into individualized recommendations based on what the patient most wants to preserve.

Key Arguments: Patients facing serious illness often need more than raw numbers; they need help interpreting risk in light of their values and priorities. Hannah Fry’s experience suggests treatment decisions can feel falsely binary when in reality there may be a spectrum of options. A 4% survival gain may be meaningful to one patient and not to another, depending on side effects, quality of life, and personal circumstances. Overdiagnosis is real: detecting more cancers does not always translate into fewer deaths. Many cancers are slow-growing or biologically indolent, meaning some people die with cancer rather than from it. The best medical communication is collaborative: clinicians should translate statistics into personalized decisions rather than overwhelm patients with data. Screening programs can create more diagnoses and more surgery without improving mortality, so outcomes must be evaluated beyond case counts.

Data Points: Age of Hannah Fry at the time of discussion: 38 - Hannah describes her cancer diagnosis and aftermath. Chance of cancer returning: 1 in 10 - Hannah is described as now cancer-free with a recurrence risk estimate. Lymphedema risk before surgery: 1 in 4 - Hannah was told this was a possible side effect of lymph node removal. 10-year survival without chemotherapy: 84% - Breast cancer patient Anne was told this estimate if no further treatment was given. 10-year survival with chemotherapy: 88% - Anne was told chemotherapy would raise survival by this amount. Absolute survival gain from chemotherapy: 4% - The episode highlights the marginal benefit discussed with the breast cancer patient. Women studied in Danish autopsy research: 77 - Researchers examined women who had died from other causes and had no known cancer diagnosis. Proportion with cancerous or precancerous changes: nearly a quarter - Found in the Danish study after double mastectomies were performed post-mortem. Estimated share of people carrying undetected cancer: 2% to 9% - Best estimates cited for people living with cancer in their body without knowing it.

Pivotal Quotes: "there's it was sort of a false choice. It really wasn't a case of either taking out everything or take out almost nothing." — Hannah Fry: Reflecting on her cancer treatment options and informed consent. "tell me what it is that you care about, tell me what it is that you want to preserve the most, and let's work out together the treatment that's right for you" — Hannah Fry: Describing the best way clinicians should communicate statistics to patients. "if you do nothing else, if we don't touch you again, if we just send you home today, your chances of surviving another 10 years are 84%. But if we give you chemotherapy, we can increase that to 88%" — Dr. Judy King: Explaining treatment trade-offs to the breast cancer patient Anne.

Implications: Listeners are encouraged to ask for absolute risks, side effects, and alternatives, not just diagnoses. For healthcare, the episode warns against overtreatment and supports shared decision-making grounded in patient values.

🔓 Sign Up for Unlimited Episode Search

About More or Less Behind the Statistics

Tim Harford and the More or Less team try to make sense of the statistics which surround us. From BBC Radio 4

View all episodes from More or Less Behind the Statistics