The grading vocabulary | Exosomal Therapy

Why a fixed vocabulary

Phrases such as promising, emerging, clinically proven and backed by science carry no fixed meaning and cannot be checked. A publication using them can shift position without anyone noticing. A fixed vocabulary makes our assessments comparable across articles and makes a change of grade an event that has to be justified.

Two scales are used. The level records the highest tier of evidence the claim has reached. The grade records our assessment of the claim given that evidence. They are separate because a claim can reach human studies and still be contested, and a claim can be plausible with no human data at all.

The level scale

Five steps, marked on each panel as a ladder. The ladder is a shape rather than a colour code, and it records only how far the evidence has travelled, not how good it is.

Levels of evidence
Level	Meaning
`Not shown`	No study at any level has demonstrated the step in question, or the step has not been tested.
`In vitro only`	Demonstrated in cells, tissue fragments or molecules outside a living organism.
`Animal models`	Demonstrated in a living non-human organism.
`Small human studies`	Demonstrated in people, in studies limited by size, blinding, control or duration.
`Larger human studies`	Demonstrated in people in studies with adequate size and defensible design.

The grade vocabulary

Seven grades. Nothing on this site may use a grade outside this set, and a claim that cannot be tested as worded receives NOT ASSESSABLE rather than being forced onto the scale.

Grades
Grade	What it requires
`NOT SUPPORTED`	Testing has been done at a level capable of showing the effect, and the effect was not seen.
`PLAUSIBLE, UNTESTED`	The mechanism is coherent with known biology, but the specific claim has not been tested in people.
`EARLY, UNREPLICATED`	Human data exist, but in small numbers, from few groups, or without independent repetition.
`SUPPORTED, LIMITED`	More than one human study points the same way, with limits on size, blinding, duration or endpoint.
`SUPPORTED, CONSISTENT`	Multiple independent human studies with defensible design point the same way.
`CONTESTED`	Human data exist and disagree, or the disagreement is about what the data mean.
`NOT ASSESSABLE`	The claim as worded cannot be tested, because a key term in it has no fixed definition.

The mandatory line

Every panel ends with a line naming what would change our assessment. It is not decorative. It states in advance the evidence we would accept, which means a later change of position can be checked against what we said we were waiting for, and a refusal to move can be challenged on the same basis.

Writing that line is also a discipline on us. If we cannot name a study that would change our view, then either the claim is not empirical or our position is not evidence-based. Both are worth discovering before publication rather than after.

What a grade is not

It is not clinical advice. A grade describes the state of published evidence for a claim, not whether a treatment is right for any individual.
It is not a safety assessment. Safety and efficacy are separate questions with separate evidence.
It is not permanent. Grades move when evidence arrives, in either direction.
It is not purchasable. No commercial arrangement of any kind can influence a grade. See the rate card.

How to challenge a grade

Send the citation. If a study exists that meets the standard set out in a panel's final line, we will read it and move the grade, and we will say what changed and when. If a study is offered that does not meet that standard, we will say why. Both exchanges are more useful published than private, so we publish them. Write to desk@exosomaltherapy.co.uk.