# Questions and answers: numbers in a treatment conversation

Based on LHD-020, r01, ARTICLE_EN.md.

## What should I ask about a percentage?

Which outcome, period, group and comparator? Ask for figures out of the same number of people over the same time. Discuss applicability with the clinician (I, IV).

## How do the two reductions differ?

In the invented example only, 20 per hundred without A and 15 with A over one year give five fewer cases per hundred: five percentage points of absolute reduction. Five out of the starting twenty is a 25% relative reduction. This is no actual treatment effect or personal forecast (II).

## Are the remaining people healthy?

Not necessarily. Eighty and eighty-five per hundred lack X over the same year; other outcomes are not described. Absence of X does not establish complete health (III).

## What if understanding and resources are limited?

Ask orally for one accessible comparison, name an important outcome and ask what remains unknown. No app, table or helper is required first. The figures prescribe neither treatment nor safe waiting; serious current danger needs appropriate urgent help (V–VIII, practice).

## What supports the explanation?

[CDC](https://www.cdc.gov/public-health-risk-communication/php/about/understanding-and-applying-risk-communication-principles.html) and the [Clear Communication Index](https://www.cdc.gov/ccindex/tool/page-20.html) concern presenting numbers; [GMC](https://www.gmc-uk.org/cdn/documents/gmc-guidance-for-doctors---decision-making-and-consent-english_pdf-84191055.pdf) supports UK professional dialogue. [Confucius](https://www.gutenberg.org/cache/epub/3330/pg3330-images.html) and [Epictetus](https://classics.mit.edu/Epictetus/epicench.html) provide no evidence of medical effects.