Glossary: evidence, GRADE and common terms

Evidence grading: A to None

The wiki uses a simplified version of the GRADE system (Grading of Recommendations, Assessment, Development and Evaluations) to indicate the strength of the scientific evidence for an effect.

---

Key concepts

Anecdotal evidence Informal personal experiences or case reports. Not usable to demonstrate cause and effect, but can generate hypotheses for formal research.

Bias Systematic error in a study that pulls results in a particular direction. Selection bias, publication bias and measurement bias are common forms. More bias means lower evidence strength and a lower grade.

Blinding Study design where participants (single-blinding) and/or researchers (double-blinding) do not know who receives treatment or placebo. Reduces bias and increases the reliability of the outcome.

Body of evidence The totality of available scientific literature on a topic. One study is rarely decisive; the coherence of multiple studies determines the strength of evidence.

Phytotherapy Use of plants or plant extracts as treatment. Falls partly under conventional pharmacology (active ingredients are measurable and dose-dependent) and partly under complementary medicine.

GRADE Internationally used system to assess the quality of scientific evidence and the strength of recommendations. Developed for human medicine, now also applied in veterinary scientific positions [2][3]. The wiki uses a simplified variant (A–D/None).

Homeopathy Treatment method based on extreme dilution and the principle that 'like cures like'. Systematic reviews in a veterinary context — including an analysis of all available randomised studies [1] — found no convincing evidence for efficacy above placebo. Scientific umbrella organisations such as EASAC [2] and the AVMA [3] reached identical conclusions.

Meta-analysis Statistical summary of multiple studies on the same topic. Increases statistical power, but is sensitive to the quality of the included studies and to publication bias.

Nocebo effect The opposite of placebo: negative expectations lead to worsening of symptoms. Relevant when assessing adverse event reports without a control group.

Nutraceutical Nutrient or dietary supplement with a presumed health-promoting or therapeutic effect (such as omega-3 or glucosamine). Legally falls between food and medicinal product; health claims are legally restricted.

Placebo effect Improvement of symptoms due to the expectation of treatment, not due to the treatment itself. In animals, the owner effect also plays a role: an owner who expects improvement may perceive and report the animal's behaviour differently — which makes clinical research in animals extra complex [1].

RCT (Randomized Controlled Trial) Randomised controlled trial. Participants are randomly allocated to a treatment group and a control group. The RCT is the gold standard for demonstrating a causal relationship. Grade A requires multiple well-conducted RCTs or a systematic review thereof [1].

Systematic review Method in which all available literature on a question is systematically searched, assessed and summarised. Higher strength of evidence than a single study, provided the methodology is transparent and the included studies are of sufficient quality.

YMYL (Your Money or Your Life) Concept from search engine quality guidelines: content that directly affects health, safety or finances. Veterinary and medical information always falls under YMYL and requires extra attention to accuracy, source citation and transparency about uncertainty.

---

How this wiki grades

Per page, the evidence grade is explicitly stated and substantiated with primary sources. No grade is 'rounded up': if evidence is lacking or contradictory, it states None or D — not C. This is not a negative judgement on a modality as a whole, but an honest representation of what the available literature currently shows.