Codeine and Breastfeeding Lancet Retraction

The Lancet Retracts the Codeine Case That Shaped 20 Years of Breastfeeding Guidance

For almost two decades, one case report has shaped how nurses, pharmacists, and prescribers talk to breastfeeding mothers about pain relief. This week, The Lancet retracted it.

The 2006 report by Gideon Koren and colleagues, “Pharmacogenetics of morphine poisoning in a breastfed neonate of a codeine-prescribed mother,” described the 2005 death of a newborn whose mother had been prescribed acetaminophen with codeine (Tylenol 3) after delivery. The authors attributed the death to the mother being a CYP2D6 ultrarapid metabolizer who passed lethal amounts of morphine to her baby through breast milk.

In its retraction notice (Vol 408, November 21, 2026), The Lancet says an investigation by the Research Integrity Office at The Hospital for Sick Children in Toronto found that the published report left out material toxicology data, including the infant’s blood codeine and acetaminophen concentrations and the identification of codeine in the infant’s stomach contents. The investigation concluded that this omission constitutes falsification, and that readers were left without the information they needed to evaluate the report’s conclusion.

In the science world, a retraction from The Lancet is a big deal. In lactation pharmacology, this one is enormous.

Why this case mattered so much

Before 2006, codeine was widely considered compatible with breastfeeding. After the Koren report, regulators moved quickly. The FDA issued warnings and, in 2017, labeling changed to say breastfeeding is not recommended during treatment with codeine or tramadol. Similar restrictions followed in the UK, Canada, and elsewhere. The case became a fixture in pharmacology lectures, nursing references, and patient handouts.

The result: a generation of postpartum mothers told to choose between pain control and breastfeeding, or handed stronger opioids instead.

The doubts were not new

Toxicologists questioned the case almost immediately. A 2008 letter in The Lancet asked whether the morphine dose delivered through milk could plausibly be fatal. In 2020, pharmacologists Jonathan Zipursky and David Juurlink published a review arguing the death could not be explained by breast milk exposure: the morphine in the milk sample was low, while the infant’s blood codeine level was far higher than breastfeeding could produce. Two Canadian journals retracted related papers in 2020 and 2021.

In January 2026, a New Yorker investigation reexamined the case in detail. Days later, The Lancet issued an Expression of Concern and referred new allegations to SickKids for a formal investigation. This week’s retraction is the outcome.

The retraction also notes that authorship and ethical concerns could not be fairly investigated because meaningful participation from Koren was unlikely to be possible.

What the omitted data suggests

Codeine found in an infant’s stomach contents is hard to explain through breastfeeding. Breast milk exposure would deliver small amounts of drug and metabolite; it would not leave codeine itself sitting in a newborn’s stomach. Experts who have reviewed the case have argued for years that the findings point to the drug being given directly to the infant, not transferred through milk.

The latest Hale’s Medications & Mothers’ Milk codeine entry already treated the case with caution, noting that one death has been “questionably attributed to codeine use in lactation,” and that “other narcotics with no active metabolite and a lower relative infant dose are preferred.”

— Hale’s Medications & Mothers’ Milk, codeine entry (L3 – Limited Data–Presumed Compatible)

Dr. Kaytlin Krutsch, coauthor of Hale’s Medications & Mothers’ Milk and Director of the InfantRisk Center, flagged the problems with this case in her August 2025 Q&A with Springer Publishing, months before the New Yorker story and the Expression of Concern.

What this means for practice (and what it doesn’t)

The retraction removes the evidence base that a mother’s breast milk alone killed a baby. It does not make codeine a preferred analgesic, and FDA labeling has not changed. For nurses, a few practical takeaways:

  • Non-opioid analgesics first. Acetaminophen and ibuprofen remain first-line for most postpartum pain.
  • If an opioid is needed, use the lowest effective dose for the shortest time, and avoid stacking other sedating medications.
  • Teach families what to watch for: unusual sleepiness, difficulty feeding or latching, limpness, or shallow or irregular breathing. Any of these warrant a call to the provider or 911.
  • Don’t default to stopping breastfeeding. Unnecessary weaning and undertreated pain carry real harms of their own.
  • Expect questions. Patients and colleagues may see headlines; follow your institution’s policy and current labeling while guidance catches up.

The bigger lesson

One sensational case report kicked off two decades of scrutiny and changed global prescribing. Follow-up case reports did appear, but largely from the same lab, and more retractions may be coming. Lactation pharmacology depends on measured data—drug levels in milk, infant plasma levels, real-world outcomes—not single dramatic anecdotes. When evidence is retracted, guidance should be revisited, and mothers deserve to hear that the science has changed.

For drug-by-drug lactation data and risk categories, clinicians rely on Hale’s Medications & Mothers’ Milk, coauthored by Thomas W. Hale and Kaytlin Krutsch.

Sources: The Lancet retraction notice (Vol 408, Nov 21, 2026); The Lancet Expression of Concern (Feb 3, 2026); Retraction Watch (Feb 4, 2026); Zipursky & Juurlink, Clinical Pharmacology & Therapeutics (2020); FDA Drug Safety Communication on codeine and tramadol (2017).

Springer Publishing Company