Aggregated answers from the case report form. Everything here is self-reported, unverified, and self-selected. It is published because the pattern matters and nobody else is collecting it — not because it can stand in for a study.
No submissions yet. The tables below fill in automatically once the first cases arrive: export them from the form service into data/submissions.csv and rebuild. The layout is complete and waiting.
The denominator that most owner discussions lack. A household reporting six cats of which two were positive is a different situation from two cats of which two were positive.
No data yet.
A "positive" means different things depending on the test and on how the sample was collected. A positive on a direct smear and a positive on a fecal-loop PCR are not the same evidence.
No data yet.
Collected by category rather than by brand. A self-selected survey cannot support a statement about a named company's product, and a site that published one would be doing something both unfair and statistically meaningless.
No data yet.
The core question. Infection has been observed to relapse as long as 20 weeks after the last dose, so a negative test taken before then does not establish clearance — it is an early reading. The split between early and late negatives below is the number worth watching.
No data yet.
The form collects no name, no email address, no location finer than nothing at all, and no free text. Responses are exported to a spreadsheet, saved as a comma-separated file, and the site is rebuilt; the tables above are computed from that file at build time rather than entered by hand, so the arithmetic is reproducible and the same input always gives the same output.
Only aggregates are published. Raw rows are never shown, because a short row — a six-cat household with two positives, a pigeon-formula source and a retest at 20 weeks — can identify a household to anyone who knows them. That is also why individual entries cannot be deleted later on request: there is nothing stored that points back to a person.
