B
Probabilities and risks in a treatment conversation
In an imagined scene, Anton hears that a risk will become a quarter smaller. He pictures twenty-five of every hundred people necessarily benefiting. Later he learns the starting figures and feels deceived. Before judging anyone's intention, however, the outcome, comparison and period need recovering. A correctly calculated percentage may convey too little meaning.
Epigraph
“When you do not know a thing, to allow that you do not know it.”
Confucius, Analects II.17, excerpt, James Legge. The passage joins acknowledging ignorance with acknowledging what is known. It does not recommend abandoning investigation.
About this article
In an imagined scene, Anton hears that a risk will become a quarter smaller. He pictures twenty-five of every hundred people necessarily benefiting. Later he learns the starting figures and feels deceived. Before judging anyone's intention, however, the outcome, comparison and period need recovering. A correctly calculated percentage may convey too little meaning.
This article helps adult patients and supporters ask about numbers while leaving room for knowledge and uncertainty. Its example is invented and describes no actual disease or treatment. It chooses no medication, calculates no personal prognosis and sets no acceptable risk for the reader. Its practical benefit is to recognise an incomplete explanation, request comparable figures and connect discussion with what matters to the person. Philosophical sources examine knowledge and action; they establish no medical effectiveness.
I. A number first needs a subject
“A ten per cent risk” does not yet say what might happen. Does it mean any unwanted effect, one serious complication, a repeat appointment or something else? One outcome must not quietly replace another. “Ten per cent of what?” is the beginning of a shared subject rather than a quibble.
Then comes time: a month, a year or another period. Ten hypothetical cases over a year cannot simply be compared with five over a month and the second option declared better. The group behind an estimate matters too. Those studied need not resemble Anton in every relevant respect. He need not determine that resemblance himself: he can ask how applicable the evidence is and where the clinician sees limitations. What, over how long and among whom restore meaning that a percentage sign cannot supply.
II. A quarter reduction does not mean twenty-five people in a hundred
Consider a teaching example only. In two imaginary comparable groups of a hundred, suppose unwanted outcome X is expected over one year in twenty people without option A and fifteen with it. These are not observations or an offer to use A. The difference is five cases per hundred people over a year: an absolute reduction of five percentage points. Five is a quarter of the starting twenty, giving a relative reduction of twenty-five per cent.
Both descriptions fit the same example. Confusion arises when “twenty-five per cent lower” becomes “twenty-five percentage points lower” or a promise of benefit to every fourth particular person. CDC recommends absolute numbers with consistent denominators. That source supports the presentation method; these figures are author-created and arithmetically checked. A useful real-life question is: “Show both figures out of the same number of people over the same period.”
III. A number's other side does not cancel its first side
In this same example, X does not occur in eighty of a hundred without A and eighty-five with it. One can describe five fewer cases or five additional people without that outcome. Both preserve the difference when the outcome and period stay fixed. “Without X” does not mean completely healthy: the example says nothing about other events.
A risk's label also does not establish its significance to someone. A rare event can matter greatly, while a frequent event needs its severity and duration explained. Different unwanted outcomes cannot simply be added like identical coins to produce a decision. Anton might particularly value retaining his way of communicating or access to assistance. That invites discussion of relevant outcomes rather than proving he must choose one option. Existing disability does not make additional harm irrelevant or allow an observer to rate an entire life on one scale.
IV. A group does not become a personal prediction
An expectation of fifteen in a hundred does not identify which people will be among the fifteen. An event does not happen fifteen per cent to an individual. This does not make statistics useless: they help compare possibilities under specified conditions. Knowing a frequency and knowing someone's future are nevertheless different forms of knowledge.
Anton can ask which of his circumstances were considered and how dependable the estimate is. The clinician may describe a range or explain that relevant evidence is limited. Acknowledging a limit does not itself establish incompetence. GMC paragraphs 21–26 concern individual discussion of benefits, harms and uncertainty. This is UK professional guidance rather than worldwide law. When information changes, ask what changed in the estimate's basis. An earlier explanation need neither become sacred nor be called deceptive simply because it was revised.
V. What remains unknown needs naming
In Analects II.17, admitting ignorance belongs to ethical learning. A person need not look informed at any price. In Anton's conversation this helps distinguish having an estimate, having limited evidence and not yet having investigated. These are different answers, none improved by invented precision.
The ancient text supplies no calculation method. Its contribution here is criticism of pretence rather than permission to replace research with sincerity. A clinician may need to clarify evidence and a patient to understand its limits. “What could be found out, and would it affect the decision?” is more useful than demanding an immediate exact percentage. Where no further answer exists, the remaining uncertainty should be named. It becomes neither safety nor catastrophe automatically, nor an obligation for the patient to find the missing study alone.
VI. A decision concerns a life as well as a calculation
In Enchiridion 1, Epictetus distinguishes our actions from external results. A limited application for Anton is that he can ask, identify priorities and participate without commanding the outcome. Stoicism's placement of the body beyond our control should not become indifference to illness or a requirement not to fear. Pain, dependence on assistance and inaccessible systems have actual consequences.
Confucian learning and Stoic work on judgement have different foundations. Contemporary clinical discussion adds professional responsibility for explanation and care. A patient need not become a perfect thinker to receive either. Clarity does not abolish vulnerability. It makes possible discussion of gains, losses and burdens significant to this person. A calculation remains part of that conversation rather than its sole judge.
VII. Understanding requires no spreadsheet of your own
With fatigue or difficulty using numbers, request one comparison through words, a drawing or a familiar accessible method. The CDC Clear Communication Index combines numerical explanation with words and visuals. This concerns organising a message rather than testing a patient's mathematical ability. A graph alone guarantees no understanding either.
Without money, an app or a family helper, one spoken sentence can begin: “I lost what this figure is being compared with.” Where substantial support is needed, agree a response method with the person; a helper may convey a question without taking over the choice. Patients need not first build a perfect diagram. Check one meaning at the end: “Have I understood that these figures concern this outcome over this period, while my result is not known in advance?” A mismatch calls for repairing communication rather than judging the questioner's worth.
VIII. When the numbers end, the question of care remains
Anton's scene need not finish with a treatment choice. A more honest ending is that he distinguishes a relative quarter from five cases in a hundred, understands the example's limits and can name his question. An actual decision needs actual evidence, an appropriate clinician and the person's circumstances. A teaching table cannot authorise starting, stopping or changing treatment.
Ask when the matter needs deciding and how further explanation can be obtained. This article sets no safe waiting period. Serious current danger requires appropriate urgent help rather than completion of a statistics lesson. Uncertainty does not release a service from conversation; understanding promises no successful outcome. Between those limits lies a meaningful action: establish what a number does and does not tell you.
What this means in practice
Choose one unclear number and ask about the outcome, period, group and comparator. Request figures out of the same number of people and the absolute difference. For a relative change, ask for the starting value. You may request descriptions of both occurrence and nonoccurrence of the same event.
Ask about applicability and remaining uncertainty. Name an outcome or burden important to you; one measure need not express your whole life. With little energy, one accessible question is enough. A record, device and family assistance are not prerequisites for requesting explanation.
The invented figures do not concern your treatment. Do not choose it from this article or postpone necessary urgent help for calculations. The exercise may produce a more precise conversation rather than proven safety, a personal forecast or obligatory agreement.
Key quotation
“Some things are in our control and others not.”
Epictetus, Enchiridion 1, opening sentence, Elizabeth Carter. Here it distinguishes participation from command of the future without denying the need for assistance.
Connections between traditions
| Comparison | Connection and boundary |
|---|---|
| Confucius II.17 / Epictetus 1 | Both limit overconfidence, but the limits of knowledge and control differ. Neither calculates clinical risk. |
| Confucian learning / contemporary research | Admitting ignorance allows investigation. Ethical honesty does not replace evidence. |
| Stoic choice / dependence on support | Participation matters without outcome control. This does not make treatment or the environment irrelevant. |
| CDC / GMC | Understandable numerical messages and individual clinical dialogue connect across distinct professional domains; they establish no common ancient doctrine. |
Sources
- CDC, Understanding and Applying Risk Communication Principles, checked 24 October 2026: main sections read; uncertainty, absolute numbers, consistent denominator and complementary frames used. Page states review on 17 September and publication on 25 September 2026.
- CDC Clear Communication Index, numeric probability: main text read; communication methods, not clinical evidence for the example.
- GMC, Decision making and consent, 21–30: actual reading covers risks, uncertainty and accessible support; UK professional guidance.
- Confucius, Analects II.17, James Legge; Epictetus, Enchiridion 1, Elizabeth Carter: primary texts in historical English translations; no critical Chinese or Greek comparison performed.
- Twenty and fifteen per hundred over a year are an author-created teaching example, not a published study. Direct access to NICE NG197 was blocked; its authority is not claimed for this article.
Tags
probability, risk, treatment, adult patient, statistics, absolute risk, relative risk, percentage point, period, group, outcome, comparison, uncertainty, evidence, explanation, accessibility, support, disability, values, burden, Confucius, Epictetus, knowledge, participation, next question
About the authors
Andrey Zarinsh is the author and founder of United Wisdom. He explores inner transformation, the unity of traditions, intercultural dialogue and contemporary spiritual discipline. Instagram: @andrey.zarinsh.
Maria Evilkut is a co-author of United Wisdom. Alongside her research, she seeks to test these ideas in practice and observe how principles of unity and mutual understanding work in real life between people and cultures. Instagram: @ilkutki.