Lancet retracts codeine-and-breastfeeding death report: Do the warnings still stand?
The journal says a SickKids investigation found that omitted toxicology data constituted falsification. The retraction withdraws a widely cited 2006 case report; Canadian product information still warns against codeine use while nursing.
In this article & details
The Lancet has retracted a 2006 report that attributed a newborn’s fatal morphine poisoning to codeine taken by his breastfeeding mother, saying a SickKids investigation found that material toxicology results had been omitted. According to the journal’s editors, the investigation concluded that the omission constituted falsification.
The withdrawal changes the standing of a case that helped shape concern about postpartum codeine prescribing. It does not itself revise drug labeling: Canadian patient information for TEVA-CODEINE still says the medicine must not be used while nursing and warns of serious harm to a breastfed infant.
The retraction notice is dated for the journal’s November 21, 2026 issue. It follows an expression of concern published online on February 3 and a new investigation by the Research Integrity Office at The Hospital for Sick Children in Toronto. Lancet retraction · Expression of concern record
The omitted results went beyond the infant’s morphine level
The editors identify three pieces of material information absent from the case report: the infant’s blood codeine concentration, the blood acetaminophen concentration and the qualitative identification of codeine in the stomach contents. The notice says omitting those findings left readers without information needed to assess the paper’s conclusion. It is that omission which the investigation classified as falsification, according to the editors. Retraction notice
The notice also records a limit to the investigation. Authorship issues and ethical concerns could not be fairly investigated because meaningful participation by Gideon Koren was unlikely to be possible. The editors do not present those unresolved matters as findings of either misconduct or exoneration. Their stated decision is to retract the case report on the basis described in the notice. The Lancet
A 2006 explanation had already faced years of challenge
The original case report, by Koren and colleagues, appeared in August 2006. Its proposed explanation linked the mother’s unusually rapid conversion of codeine into morphine with the infant’s exposure through breast milk. That explanation became the subject of a prolonged scientific dispute.
In a 2020 pharmacology review, Jonathan Zipursky and David Juurlink challenged whether the reported death could plausibly be explained by breastfeeding exposure. They pointed to the small quantities of opioids transferred into milk, the pharmacokinetic conditions needed for substantial accumulation in a newborn and a previously unreported, markedly elevated codeine concentration in postmortem blood. Those were the reviewers’ arguments against the case’s proposed mechanism, rather than a new experiment establishing what happened to the infant.
Related articles had already been withdrawn by Canadian journals. In a November 2020 joint retraction statement, Canadian Family Physician and the Canadian Pharmacists Journal said two independent experts had concluded that their columns met the publication-ethics criterion for unreliable findings. The editors said they had reviewed the original report, the critique, correspondence from a parent and the original authors’ responses before seeking that independent assessment.
The two withdrawn columns appeared in the pharmacy journal in 2006 and the family-medicine journal in 2007. The new Lancet notice concerns the original case report, completing a separate journal’s review process rather than announcing the first withdrawal of reporting about this case. Canadian journals’ statement · Lancet notice
Later Ontario research examined outcomes beyond one case
A 2023 Ontario population study in The BMJ compared 85,675 mothers who filled an opioid prescription shortly after discharge following delivery with an equal number who did not. Infant hospital admission within 30 days was 3.5% in both groups. Emergency-department visits were slightly more frequent in the prescription group, but the study found no infant deaths and no increase in the other measured adverse outcomes.
The findings have limits. The study covered several opioids and did not establish how much medicine each mother actually took. Researchers also lacked breastfeeding-status data; less breastfeeding among opioid recipients could have obscured an exposure-related risk. The study therefore supplies broader outcome evidence without proving that every opioid regimen is safe during breastfeeding. Study methods and limitations
Retraction and prescribing advice remain separate questions
The Canadian TEVA-CODEINE information lists breastfeeding among the situations in which the medicine must not be used. It also tells families to seek immediate medical care if an exposed infant is excessively sleepy, has feeding or breathing difficulties, or becomes floppy. Those product instructions remain distinct from the journal’s judgment about the reliability of the 2006 report. Canadian product information
The October 15, 2026 LactMed review takes a more qualified view of the evidence than a blanket statement that one discredited report proves safety or danger. It notes that many organizations prefer alternatives or advise avoiding codeine, while discussing the challenged case and reassuring population studies. It also cautions that maternal oral opioid use can cause infant drowsiness and, rarely, more serious central nervous system depression, with newborns particularly sensitive.
LactMed’s assessment predates this Lancet retraction. The publication record has now changed again: the original case report has been withdrawn, while clinical references and product instructions must be assessed on their own evidence and current wording. The retraction is a finding about the published report’s reliability, not a new authorization to use codeine while breastfeeding. Retraction · Canadian product information · LactMed
References
- Retraction: Pharmacogenetics of morphine poisoning in a breastfed neonate of a codeine-prescribed mother — November 2026 — The Editors of The Lancet
- Expression of Concern: Pharmacogenetics of morphine poisoning in a breastfed neonate of a codeine-prescribed mother — February 3, 2026 — The Lancet / PubMed
- RETRACTED: Pharmacogenetics of morphine poisoning in a breastfed neonate of a codeine-prescribed mother — 2006 case report — The Lancet / PubMed
- Risks of maternal codeine intake in breastfed infants: a joint statement of retraction — November 2020 — Canadian Family Physician / Canadian Pharmacists Journal
- The implausibility of neonatal opioid toxicity from breastfeeding — 2020 — Clinical Pharmacology & Therapeutics
- Maternal opioid treatment after delivery and risk of adverse infant outcomes: population based cohort study — March 15, 2023 — The BMJ
- TEVA-CODEINE patient information — pregnancy, nursing, labour and delivery warnings — Health Canada Drug and Health Product Register
- Codeine — Drugs and Lactation Database (LactMed), revised October 15, 2026 — U.S. National Library of Medicine / LactMed