Name the quantity before calculating it
Decide whether you are describing animals affected, premises affected, positive samples or reported events. Keep the unit in the column heading and the written explanation. A table headed “cases” without a definition can mix incompatible records.
Specify the period and population. “Positive among samples submitted to this laboratory during this interval” is narrower than “prevalence in all cattle in the region.” The narrower statement may be the one the data actually support. Choosing more dramatic wording does not expand the evidence.
Inspect the denominator
Ask which animals or holdings could have entered the count, which were actually assessed and which were omitted. Record known eligibility rules and gaps. A denominator should describe the same population, period and unit as the numerator.
| Before publishing | Question to answer |
|---|---|
| A percentage | What exactly is above and below the division line? |
| A time trend | Did selection, testing or definitions change? |
| A regional comparison | Are the populations and detection opportunities comparable? |
| A zero | Was absence assessed, or is information missing? |
| A probability claim | What method supports that interpretation? |
Keep calculated results at a precision justified by the inputs. A percentage with several decimal places is not more reliable when group size or case classification is uncertain.
Work through a submitted-sample example
Suppose, as a fictional arithmetic example, a laboratory receives twenty samples because owners already suspect a problem, and eight meet a specified result category. Eight of twenty is forty percent of those submissions. It is not automatically forty percent of animals in the source region.
The selection process matters even when the arithmetic is correct. If the next month includes routine screening as well as concern-driven submissions, comparing the two positive proportions requires that changed selection to be explained. Do not label the difference an increase or decrease in regional prevalence without an appropriate method.
Preserve missing and negative information separately
An empty record may mean no submission, no report, a delayed update or an unavailable field. A negative finding means an assessment occurred with a stated result and scope. Keep those states distinct in both the dataset and the prose.
If an official source supplies occurrence categories or definitions, retain them with the extraction. Avoid collapsing “not reported,” “not assessed” and a defined absence category into one reassuring green symbol. A reader should be able to see what was actually established.
Write the conclusion at the supported scale
State the source, unit, population and interval beside the finding. Explain selection limits in the same paragraph rather than hiding them in a distant note. When the necessary denominator or method is unavailable, publish a bounded count or description and leave the prevalence question unresolved.
Use historical outbreak-data guidance for dates and event duplication, and surveillance interpretation for broader context. A useful public reference can describe uncertainty clearly without inventing either a precise risk or a claim of complete safety.
Test the sentence by replacing “cases” with its real unit
Before publication, substitute the actual counted object into the headline. “Eight positive submitted samples” is clearer than “eight cases” when no event or animal definition has been established. If several samples came from the same animal, that distinction can change both the count and the denominator needed for a proportion.
In a second fictional dataset, three reports concern one premises and two are amendments. A summary initially calls them three affected farms. Linking the reports to their actual premises and version relationships preserves the document count without inflating the premises count. The correction does not alter the underlying findings; it repairs the unit being reported.
For a trend, write down what changed in the observation system before interpreting the line. Wider eligibility, a new submission route or a revised definition can affect totals. Keep these explanations alongside the series so readers do not have to infer whether the reporting process stayed constant.
Frequently asked questions
Is a correct percentage automatically a prevalence estimate?
The arithmetic can be correct while the population claim is unsupported. Identify how animals or premises entered the dataset and whether the denominator represents the population described. Publish the narrower supported quantity where necessary.
Can a missing report be displayed as zero?
Keep missing, delayed, unassessed and measured negative states distinct. A zero should have a defined observation basis. Otherwise a reassuring display can arise from an information gap.
What if no suitable denominator is available?
Describe the bounded count with its source, unit and period and explain the selection limit. Leave the broader prevalence or probability question unresolved. Extra decimal places, a map color or a confident title cannot supply the missing population information.
Cite this reference
LivestockDefense. “Avoid turning disease reports into unsupported prevalence claims.” Updated 2026-11-03. https://livestockdefense.com/guides/avoiding-unsupported-disease-prevalence-claims/
Add the date you accessed the page. This identifies the current reference; keep a permitted saved copy if your work depends on its exact wording. A citation alone does not establish support for a claim.
How to cite a page and its revision ↗