- Design
- Disproportionality analysis of FAERS and JADER with systematic review
- Population
- Spontaneous adverse event reports for ranitidine and famotidine; comparative safety studies
- Primary outcome
- Cancer reporting signals
- Effect
- Signals attenuated vs famotidine and pre-recall; JADER PRR 3.54 to 0.92
Ranitidine was withdrawn worldwide in 2020 after N-nitrosodimethylamine (NDMA), a probable carcinogen, was found in products. This study, published in Drug Safety on 10 October, asked whether the cancer signals that followed in adverse event databases reflect a real effect.
In the US FDA database, cancer reports for ranitidine surged in 2020–21 to 140,985 — over two-thirds of all cancer reports in that period. Most came from consumers and clustered where regulatory action and litigation were concentrated. Unrestricted disproportionality analysis produced large signals, especially for gastrointestinal and genitourinary cancers. These signals largely disappeared when ranitidine was compared with famotidine, a similar drug without the contamination, and when analysis was restricted to before the recall. In the Japanese database, a signal among reports naming ranitidine as suspect (PRR 3.54) vanished when all reports involving the drug were included (PRR 0.92). The accompanying systematic review found most large cohort studies showed no increase in overall or site-specific cancer.
The authors attribute the signals mainly to stimulated reporting and protopathic bias — the drug was taken for symptoms of an as-yet undiagnosed cancer. Spontaneous reports are for generating hypotheses; this is a clear example of why they cannot measure risk.
- A disproportionality signal from spontaneous reports is a hypothesis, not a measure of risk.
- Ask whether a signal persists against an active comparator and outside periods of publicity or litigation.
- Consider protopathic bias when a drug for early symptoms (dyspepsia, pain) seems linked to a cancer that causes those symptoms.
- Patients who took ranitidine can be told large cohort studies have mostly not shown increased cancer risk.
Why it matters
It shows how publicity can manufacture a statistical safety signal, which matters every time a new pharmacovigilance alarm reaches patients.
Don't overread it
This does not show NDMA contamination was harmless; it shows spontaneous reports could not measure the risk.
The statistics, in plain English
A proportional reporting ratio (PRR) compares how often an event is reported with one drug against all others; it reflects reporting behaviour as much as risk. A PRR of 3.54 that falls to 0.92 when reporting rules change shows how fragile such signals can be.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for clinical pharmacology, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free