EVIDENCE LIBRARY
How Risky Is One Ecstasy Pill? Why the Denominator Matters
A review estimated nonfatal and fatal ecstasy risks using national monitoring data, but its assumptions show why apparent precision must be interpreted cautiously.
Statements that a drug is “safe” or “dangerous” are incomplete unless they identify the outcome and denominator. Risk per user, per year, per session, and per pill are not interchangeable. A 2020 review attempted to estimate acute ecstasy-related risk using national incident and exposure data.
What did the authors estimate?
For nonfatal incidents, the researchers combined Dutch monitoring data with estimates of national ecstasy use. They calculated an approximate risk of one moderate-to-severe acute incident per 900 pills, or 0.11%. Their comparable estimates were one per 1,600 cocaine doses and one per 95 gamma-hydroxybutyrate doses.
For deaths, they used Great Britain mortality data and estimated a fatal risk attributed to MDMA alone of about one per 33,000 pills. Expressed per user, the estimated range was 0.01% to 0.06%, depending on the assumptions and jurisdiction.
Where did the numbers come from?
The nonfatal estimate was not derived from a prospective cohort following every pill. It used a Dutch drug-incident monitoring program, adjusted for incomplete national coverage, assumed an average of 1.2 pills per session and four sessions per year, and combined those figures with prevalence estimates.
Each multiplier introduces uncertainty. Underreporting, tourists, severity classification, changing pill strength, repeat users, and polydrug exposure can all alter the numerator or denominator.
Why does “MDMA-related” need definition?
A death certificate or incident record may list MDMA while alcohol, cocaine, amphetamine, or another drug is also present. The review separated deaths involving MDMA alone from broader MDMA-related deaths where possible. The difference was substantial, especially across regions.
That distinction does not prove that a case with only MDMA detected was pharmacologically simple. Nor does co-detection establish that every substance materially contributed. It does show why “mentioned,” “detected,” and “caused” should not be treated as synonyms.
What can the estimate support?
The paper supports two positions at once. Severe incidents were uncommon relative to the enormous number of estimated use occasions, yet ecstasy was not harmless. Hyperthermia and hyponatremia were among the severe clinical patterns, and fatal cases occurred.
It does not support telling an individual that a specific pill carries a known 0.11% risk. Tablet dose and composition vary, individual susceptibility differs, and setting and co-use matter. The estimate describes a monitored population over a defined period.
Questions for interpreting any risk statistic
- Is the denominator users, sessions, doses, or pills?
- How were incidents captured, and what proportion was missed?
- Does the numerator include co-use?
- How was severity defined?
- Were pill strength and drug-market changes considered?
- Is the estimate transferable to another country or year?
Risk estimates are useful when their assumptions remain visible. Removing the denominator or uncertainty can make a population estimate appear more precise—and more personal—than the evidence permits.
Scientific source
van Amsterdam J, Pennings E, van den Brink W. “Fatal and non-fatal health incidents related to recreational ecstasy use.” Journal of Psychopharmacology (2020). PubMed PMID 31909673; DOI 10.1177/0269881119897559.