Rules
Part of Health alert reporting guide for daily news
Health news alert problems: how to read notice limits
Health news alert problems explained: read a notice’s category, evidence, scope, recall status and current instructions without overstating its claims.
Health news alert problems often start when a headline makes a notice sound broader or more conclusive than it is. This explainer focuses on reading a notice’s category, evidence, scope and current instructions, so coverage does not overstate what the source establishes.
| Newsroom case | Wording to preserve | Check before publication |
|---|---|---|
| Official signal | The issuer's category for the notice | Do not upgrade an alert, investigation, advisory or recall |
| Evidence claim | Data type and review stage | A report or sequence does not by itself establish cause |
| Risk figure | Percentage with count, denominator and periods | Keep the population and comparison period attached |
| Product notice | Exact product identifiers and scope | Confirm the affected lot, model, size and expiration window |
| Recall status | Initiating party and any regulator-assigned class | Publicity is not a recall classification |
| Reader instruction | Audience, conditions and current action | Do not turn clinician guidance into a patient order |
| Changed advice | The latest instruction across every prominent surface | Replace outdated action text and add a dated note |
| Safety claim | Evidence, population, place and time | Do not promise zero risk |
How to read the category in a health notice
Start with the issuer's own category for the event: alert, report, investigation, advisory or recall. Agencies use these terms with different weights, so substituting one for another can make a routine surveillance flag sound like an emergency. A forceful headline does not change the category the source used.
A short headline can drop qualifiers even when its sentence remains grammatical. To see what the notice actually establishes, check its precise scope, evidence, intended audience and current action. Those details help distinguish a source's findings from a headline's shorthand.
When officials are investigating a cluster, the investigation is the known fact. Use crisis language only if a source applies it to this event on the record. A more urgent label in coverage does not change the source's chosen category.
What an adverse-event report can establish
An adverse-event report, complaint or case report establishes that someone recorded an event, not what caused it. Check the data type and its place in review; sequence alone does not establish causation. If timing is all that is known, coverage should say the event was reported after exposure.
Look for whether the evidence includes a comparison group, laboratory confirmation, exposure evidence or a regulator's causality finding. Coverage should state the evidence type and review stage in terms consistent with health evidence and alert types. If the record establishes sequence alone, it does not imply more.
The National Library of Medicine advises readers to check who runs a health site, its purpose, its original source, its review process, its update date and its privacy practices. The MedlinePlus guide to evaluating health information helps trace a claim to its evidence rather than relying on a polished presentation.
Read percentages against their baseline
A rise in a rare outcome may involve only a few cases, while a case count without a denominator gives no sense of scale. Look for counts and denominators when the source supplies them, along with the baseline period and the follow-up period.
Keep individual risk distinct from a population estimate. Do not read a result as one in X unless the study supports that expression. The numerator, denominator, population and comparison period belong beside a percentage; if the source leaves one out, the gap should be stated rather than filled by inference.
A relative increase is not enough to judge importance: readers also need the baseline, absolute difference, measured outcome, follow-up period and uncertainty. Without those details, the percentage cannot be read in context. The same denominator shortcut matters in a related reporting example.
Check which products a notice covers
A headline may name a brand when the notice applies only to one lot, model, size or expiration window. Compare the notice with product labels and check each exact identifier it specifies; note when a field does not apply or is not provided.
Product identifiers to verify
- Brand and product name as printed
- Full lot or model code
- Size, strength, packaging unit and count
- Expiration or use-by window
- Distribution areastates, stores, dates
- Action requestedreturn, quarantine, call
| Field | What goes in it | Why it blocks publication |
|---|---|---|
| Lot or batch code | Exact code shown on the notice and package | A different code may be outside the recall. |
| Model | Exact model or serial range, when applicable | Similar models may not be affected. |
| Size | Exact size or package variant, when applicable | A brand can include unaffected variants. |
| Expiration window | Exact range listed in the notice, when applicable | Products outside the stated window may not be included. |
If readers could confuse covered and unaffected goods, the opening line should make clear that not all products are included. Check the publication-blocking fields in the health alert reporting checklist. Never shorten an alphanumeric identifier: a changed lot or model code can point to the wrong product.
Tell visibility apart from recall status
A widely shared company notice remains a company notice, and a market withdrawal is not automatically a recall. Check who initiated the action and when. A risk class applies only when the regulator has listed one; neither a notice's reach nor its wording establishes a class.
FDA assigns three risk classes to drug recalls. Class I signals a reasonable probability of serious adverse health consequences or death. Class II means possible temporary or medically reversible consequences, or a remote probability of serious consequences. Class III means serious adverse health consequences are not likely.
Use a class only when FDA assigns it, not based on a company's wording or a notice's visibility. FDA's general recall framework, including recall classifications, appears in 21 CFR Part 7. Check the regulator's record for the classification.
For food, section 423 of the Federal Food, Drug, and Cosmetic Act, added by the Food Safety Modernization Act, gives FDA mandatory-recall authority in specified circumstances. That authority is distinct from a company's voluntary recall, so describe who initiated the action rather than blurring the two.
FDA generally uses public notification for widely distributed products or serious hazards, but not every recall appears on its website or in news coverage. Its consumer guide to drug recalls explains the three classes and tells consumers not to stop a medicine without first speaking to a health professional.
Check the issuer's own record: FDA's Enforcement Report lists recall entries, while MedWatch is FDA's safety-information and adverse-event reporting program. CDC's Health Alert Network (HAN) distributes public-health alerts; it is not an FDA recall record. Match the record to the product and jurisdiction before describing the notice.
Other US issuers include CDC, USDA's Food Safety and Inspection Service (FSIS), the Consumer Product Safety Commission (CPSC) and state health departments, including the California Department of Public Health and the New York State Department of Health. International sources include WHO Disease Outbreak News, EMA safety communications, the UK's MHRA Drug Safety Update and Health Canada's Recalls and Safety Alerts.
Check who each instruction addresses
Keep directions attached to the people they address. A clinician's monitoring instruction is not necessarily a patient order: changing the audience changes who is being told to act and what action is required. Look for separate directions for product users, clinicians and facilities.
Check the conditions attached to testing, monitoring or treatment. When a condition is dropped, a qualified direction can read like a general instruction. A recall notice is not independent treatment advice from a newsroom, so check the responsible authority's qualifications.
Check whether the action has changed
When an agency changes the affected area or instruction, older wording may remain visible in a headline, search result or social card even after the body has been updated. That can leave readers with an outdated action. Check whether the most prominent wording matches the latest notice.
The headline, summary, action line, captions and body should carry the current instruction, with a dated note explaining what changed. The earlier version can remain in the editorial record without leaving a superseded instruction where readers first land.
The sentence that tells readers what to do is especially important, followed by the headline, summary, product box and distribution text. Readers may encounter these surfaces before reaching a correction, so they need to carry the current instruction.
Keep safety claims within their limits
A lack of confirmed cases does not prove a product or situation is safe, and a detection in one setting does not show danger everywhere. Check which evidence, population, place and time a statement describes; coverage should distinguish what officials found from what remains unknown.
A protective action does not promise zero risk. A result or warning may apply to a defined group, location or period rather than everyone. Keeping those boundaries visible prevents a qualified finding from becoming a universal claim.
When coverage says a product is linked to an outcome, check whether the basis is epidemiology, traceback, laboratory testing or a regulator's finding. Where the only established fact is sequence, the wording should not imply a causal conclusion that the evidence does not establish.
