How we research
Last updated: 14 August 2026
How a question becomes a piece of guidance, step by step.
Step 1 — Frame the decision, not the topic
We start from the choice a reader is actually facing (“should I try this?”, “is this worth the money?”, “what do I ask my doctor?”) rather than from a keyword. The framing determines what evidence is relevant.
Step 2 — Search for the best available evidence
We look first for systematic reviews, meta-analyses and clinical guidance from recognised bodies, then for randomised controlled trials, then for observational studies. We look for contradicting evidence deliberately, not only for evidence that supports a convenient conclusion.
Step 3 — Assess quality, not just conclusions
We look at study size, design, population, duration, comparators, outcome measures, conflicts of interest and whether the finding has been replicated. A small, short, industry-funded trial is not treated as equivalent to a large independent one.
Step 4 — Judge relevance
Evidence drawn from populations, doses or conditions unlike the reader’s is flagged as such. We are explicit when we are extrapolating.
Step 5 — State the uncertainty
Every conclusion carries an honest description of how confident it is and why. Where the honest answer is “we do not know”, that is what we publish.
Step 6 — Connect it to real constraints
Cost, effort, availability, side effects, and how long a change takes to show an effect are part of the guidance, not an afterthought.
Step 7 — Editorial check before publication
A human editor checks that every substantive claim traces to a cited source, that uncertainty is represented fairly, that no commercial relationship has shaped the conclusion, and that the medical disclaimer is present and appropriate.
Step 8 — Revisit
Evidence changes. Guidance is dated and revisited, and material changes are logged under our corrections policy.
Current limitations
We do not currently employ in-house clinicians, and we do not yet run a formal external medical review step. Our medical review policy describes exactly what review does and does not happen today.
