How Public Health Trace collects, connects, updates, and corrects public records
Public Health Trace doesn't generate its own recall, outbreak, or surveillance data -- every record traces back to a government source. What Public Health Trace actually does is collect that data on a schedule, connect related records to each other, keep them current, and fix mistakes when they're found. This guide walks through that full lifecycle in one place; see methodology and corrections policy for the complete technical detail behind each step.
1. Collect
Each registered government source -- a CDC, FDA, USDA, or EPA feed -- is fetched on its own schedule by a dedicated importer built for that specific source's format. This matters because agencies publish on very different schedules: FDA's Enforcement Report database is published weekly, while an active CDC outbreak investigation page can change at any time as new case counts come in. Public Health Trace records what it fetched, when, and whether the fetch succeeded, so every source's freshness can be tracked independently rather than assumed.
2. Connect
A raw record on its own is just one agency's data point. Public Health Trace links it to shared company, brand, food, and disease records, so you can see every recall tied to a company, or every outbreak tied to a food, even when they came from different source feeds. Separately, some recall and outbreak pages show each other as related -- but not every shown connection is equally certain. See how recalls and outbreaks are connected for the real distinction between a confirmed, source-published relationship and a text-based match that's a starting point for your own research, not a verified fact.
3. Update
Public Health Trace re-checks each source on a tracked schedule, not continuously -- for a source with a confirmed live feed, an internal health check flags it if a successful import hasn't completed within 48 hours. A record itself can also be marked provisional, meaning the agency that published it says it may still revise the figures (most commonly, case counts for a recent outbreak CDC is still investigating). Because imports run on a schedule rather than in real time, there can be a short lag between an agency publishing something and it appearing here -- which is also exactly why every record links back to the agency's own current page.
4. Correct
If something on Public Health Trace looks wrong, outdated, or missing, we want to know. Reports go through our contact form; we first check whether the issue is in how Public Health Trace imported or displayed the data (which we fix directly) or whether it's already present in the original government record (which we can't alter, but will note and point you to the agency's own correction channel for). See corrections policy for the full process and what to expect.
What this guide is, and isn't
This is Public Health Trace's own description of its own process -- it is not a claim about how any government agency runs its own data systems, and none of the steps above involve medical judgment of any kind. See our editorial policy for how page text itself (titles, summaries) is generated, and our medical disclaimer for what this site is not.
Limitations: not every source has a machine-readable feed yet, and some fields (like recall lot codes) aren't captured by every importer -- see methodology's known limitations for the current, specific list, which changes as importers are added or improved.