1. Start with the primary record
Use an original government notice, public agency statement, research paper or issuer’s call where available. Keep the source URL, its date and the evidence review date visible. A document can establish the statement it contains without independently proving the statement’s outcome. An official announcement must retain attribution to the issuing institution.
2. Separate financing stages
Announced, committed, disbursed and spent are different stages. A procurement notice is not an award. An award is not payment. A payment is not proof of service delivery. Amounts keep their original currency and period. The monitor groups displayed amounts by currency and stage; it does not construct a national total from non-comparable records.
3. Preserve study scope
Research findings belong to the participants, methods, setting and reporting period described in the original paper. Associations should not become causal claims. A diagnostic performance measure needs its denominator, comparator and uncertainty. Evidence from one African setting does not establish Kenyan approval, availability or national prevalence. The complete methods and limitations remain in the linked study.
4. Make reuse visible
Research briefings contain licensed excerpts, with author, journal, DOI and licence attribution. The source’s wording is identified as an excerpt. Selection, layout and HealthSignal’s contextual notes are identified separately. Illustrative stock photographs are captioned as illustrations and are not evidence of a reported event or the identities of study participants.
5. Connect the record
Contextual article links are supplemented by topic, country and desk connections. Related reporting provides a route to further reading; similarity of topic is not proof that study results are directly comparable. Source records connect the underlying evidence back to reporting. Readers can use those links to trace the basis of an article instead of relying on a detached headline.
6. Review rather than forecast without data
An evidence review records the source available at a stated date. It does not predict an outbreak or certify a future funding gap. New notices, implementation records, corrections and retractions require dated updates. Questions and uncertainty remain visible when the available source cannot settle them.