A citation is not proof by itself. The useful question is whether the cited source supports the exact statement being made — in the same population, context and level of certainty.
Claim-to-source checking
Clinora traces substantive medical claims back to the reference material supplied for the project. We look for wording that goes beyond the source, removes a condition, changes the population being discussed or turns an association into something more definite than the evidence allows.
Where only part of a sentence is supported, the comment identifies that distinction. The aim is to preserve the intended message where possible while bringing the wording back into line with the evidence.
What we check
Substantiation
Does the source support the specific medical claim, not merely the general topic?
Population and context
Has evidence from a particular age group, disease stage, study setting or patient population been presented too broadly?
Numbers and statistics
Percentages, rates, ranges, denominators, timeframes and study figures are checked against the source where included in scope.
Strength of wording
We flag language that implies certainty, causation, superiority or generalisability beyond what the source actually demonstrates.
Reference details
Citation information, links and reference numbering can be checked for completeness and alignment with the agreed style.
Currentness
Sources that appear dated, superseded or unsuitable for the claim can be flagged for the client team to update.
Source quality depends on the claim
There is no single reference hierarchy that fits every sentence. A product statement may need approved product information. A treatment recommendation may be better anchored to current clinical guidance. A specific trial result should be checked against the trial publication itself. A broad evidence summary may be better supported by a high-quality systematic review.
For that reason, Clinora does not automatically replace the client's reference pack with different sources during a standard fact-check. We first assess whether the supplied source does the job it is being asked to do. If it does not, that is flagged. New literature research can be scoped separately.
Numerical claims get their own pass
Numbers are easy to trust because they look precise. They are also easy to detach from their denominator, timeframe or study population. Where numerical verification is part of the scope, figures are checked back to the source and read in context rather than compared as isolated digits.
This includes percentages, prevalence figures, risk statements, study outcomes, ranges, frequencies and other quantitative claims used in the content.
Useful for older and inherited content
Reference review is particularly useful when a team inherits a content library with uneven sourcing. It can also help before a website migration, after several rounds of editing, or when a content set has accumulated references over a long period without a structured review.
Medical fact-checking is not a substitute for legal, regulatory, pharmacovigilance or formal promotional approval. It also does not include a systematic literature review unless that work is separately agreed.
How reference queries are written
We avoid vague comments such as “needs a better source” where the problem can be stated more precisely. A useful query says what is missing: for example, the reference supports adults but the copy says “people” generally, the cited paper reports a secondary endpoint rather than the primary claim, or the percentage in the copy does not match the study figure.
Fact-checking can be commissioned on its own or combined with a wider medical content review or healthcare website review.
Frequently asked questions
Can you work from references already selected by our team?
Yes. That is the normal starting point. We assess whether those sources substantiate the wording they are attached to and flag gaps rather than silently changing the evidence base.
Can you find new references?
Targeted source research can be scoped separately. A standard reference review focuses on the content and source pack supplied by the client.
Do you check statistics and study results?
Yes, when the relevant source material is available and numerical verification is part of the agreed scope.
Can you review references across a whole website?
Yes. For larger sites, a page inventory and reference tracker help keep claims, citations and recurring source issues organised.