How TortIntel works
TortIntel shows where each active mass tort stands and what changed this week, using only public sources you can check yourself.
Where the data comes from
- Federal dockets. New cases filed in each tort's coordinating court that name its defendants, from CourtListener's index of federal PACER dockets. Checked every six hours.
- FDA recalls. The FDA Enforcement Report, limited to recalls by the tort's own defendants. A compounding pharmacy's recall is not an Ozempic development, so it is left out. Checked every six hours.
- FDA drug labels. We watch each drug's prescribing information and flag when its boxed warning or warnings section changes. Checked daily.
- Federal Register. Rules and proposed rules from EPA, FDA, CPSC and related agencies whose titles name the product or exposure. Checked daily.
- PubMed. New studies on the specific exposure and injury. Checked daily.
- News. Articles from established news and legal outlets only. Law firm pages and lead-generation sites are excluded. Checked every two hours.
How an item gets attached to a tort
An item is attached to a tort only when it names that tort's product, defendants or exposure, and news headlines must also describe a legal, regulatory or scientific development. Each news item is then checked a second time by a language model that must answer yes. When an item fits no tort, it is dropped rather than filed somewhere close. That is why some tort pages are quiet: a quiet page means a quiet docket, not a broken feed.
Case events such as trial dates, rulings and settlements are also added by hand after they are verified against a court record or primary source.
Who runs TortIntel and how items are checked
TortIntel is maintained by Jacob Malherbe, founder and CEO of Mass Tort Ad Agency (a Meta Business Partner and Inc. 5000 honoree) and author of A Lawyer's Guide to Mass Torts. He has run Meta advertising for mass tort campaigns since 2015 and developed the Jacob Score.
Automated items are published only after they pass the matching rules above. Case events such as rulings, trial dates and settlements are entered by hand after they are checked against a court record or primary source, and stages and scores are reviewed as dockets move. To report an error, email jacob@tortintel.ai.
Stages
- Individual filings. Cases are being filed one by one. No coordinated proceeding yet.
- Coordinated (MDL or JCCP). Cases are centralized before one judge for pretrial work.
- Bellwether trials. Test trials are set or under way.
- Settlement. A global or large-scale settlement is proposed, approved or paying.
- Resolved or winding down. Most claims are resolved, dismissed or closed to new filings.
Stages are set editorially and reviewed as dockets move.
The Jacob Score
The Jacob Score rates the strength of a litigation from 0 to 100. It combines five dimensions, each rated from 1 to 5:
- Science. Study quality, size of the risk, FDA or agency action, and how well the causation evidence is likely to survive Daubert.
- Momentum. Coordination status, filing trend and how quickly the docket is moving.
- Campaign. Whether claims can still be filed, the size of the qualifying population and defendant solvency.
- Timeline. How close the litigation is to bellwether trials or a resolution.
- Verdicts. Plaintiff wins, bellwether results and settlement status.
A high score means a litigation is proven, not that it is the best place to start a campaign. Early torts often score in the 30s and 40s while advertising costs are still low.
Limits
CourtListener indexes the federal dockets it can see, so filing counts show the direction of a docket rather than an official case total. Cases filed directly into an MDL's master docket, or in state court, may not appear. For official counts, use the JPML's monthly pending MDL report.
TortIntel is editorial research for plaintiff-side professionals. It is not legal advice and must not be used to evaluate an individual case.