Methods · How this site reads research
How Hussain evaluates health evidence
Every article, guide and Decision Brief on this site follows the same working rules. This page states them plainly, so you can hold any page here to them. Nothing below is a credential and nothing below is a promise of certainty: it is a description of how the reading is done.
Where claims come from
Sources are used in a deliberate order. When a stronger source contradicts a weaker one, the stronger source wins and the page says so.
- Systematic reviews and meta-analyses of human trials, read first because they summarise many studies rather than one.
- Randomised controlled trials in humans, weighed by size, duration and who was actually studied.
- Observational and cohort studies, used for patterns and risk associations, never for proof of cause.
- Mechanistic, animal and cell work, always labelled as such. It explains how something could work, not whether it does in people.
- Official UK sources, such as NHS, NICE and MHRA publications, for what is available, what is approved and how services actually work.
Single small studies, conference abstracts and preprints are treated as leads to follow, not conclusions to publish.
How studies are weighed
- Size and duration. A six-week study of forty people cannot answer a lifetime question.
- Who was studied. Results in young athletes, in one sex, or in another health system may not transfer to the person asking.
- Funding and interests. Where a cited study was funded by a party with a stake in the answer, and that stake bears on the claim, the page says so.
- Replication. One striking result is a question; the same result found independently is an answer taking shape.
- Absolute against relative numbers. A halved risk can mean two cases per thousand instead of four. Pages state the absolute difference where the source reports it.
How uncertainty is stated
Each substantial page separates what is established, what is likely, what remains uncertain and what is simply unknown, in words rather than scores. This site does not award formal evidence grades and does not invent certainty to make a cleaner headline. When the honest answer is that the evidence does not say, the page says that.
Where an older page reads more confident than its sources support, it belongs to an earlier era of this archive and is queued for refinement. Factual corrections are welcome through the contact route on the About section.
What this method cannot do
This method reads research. It does not examine patients, it does not diagnose, it does not recommend treatments and it does not interpret your personal results.
The research prepares the conversation; clinical decisions stay with you and your clinician.
Where to see the method applied
The same rules run through the whole archive and through every Clarity Brief. The Sample Clarity Brief shows the structure the method produces for one agreed question, and the Research archive shows it applied at scale.