Editorial policy
Our editorial question is whether the documented preparation and the cited evidence answer the same question. We keep the steps visible rather than combining names, numbers and outcomes into a product score.
Keep the identity attached
An organism’s genus, species and strain are recorded at the level the source provides. An absent strain code is not supplied from a related paper. A study preparation retains its original formula and participant context; a shared seller or species does not turn that study into a trial of a current retail product.
Keep the measurement attached
Mass, CFU and AFU remain the units stated in the relevant record. The articles do not derive an exchange rate, a common potency scale or a personal amount from those declarations. Public standards information is described within the portion actually accessed, and a draft regulator document remains identified as a draft.
Keep the method and outcome attached
The guides distinguish attributed consumer reports, controlled research and biological explanations. The WBF preparation, trial population and glucose-related endpoints are not relabeled as retail-product outcomes or evidence about prescription side effects. Editorial checks compare prose with its cited record; they are not clinician review, independent laboratory testing or personal treatment clearance.
Show how the record was available
Each reference retains its name, address, access description and checked date. Those dates describe source work, not a promise that an offer is unchanged today. The Culturelle refund discussion relies on indexed official-page text reported crawled two weeks earlier after direct access failed. The failed request is not treated as policy evidence, and the recovered text is not described as a tested refund.
Leave a discrepancy visible
Missing strain identifiers, unknown counts and the CoreAge six-bottle price arithmetic remain explicit. Unresolved details are neither filled from another product nor treated as proof of ineffectiveness. A more complete label can improve a question without resolving the individual care decision.