Risks of maternal codeine intake in breastfed infants: a joint statement of retraction from Canadian Family Physician and the Canadian Pharmacists Journal

This paper is jointly published in Canadian Family Physician (CFP) and the Canadian Pharmacists Journal (CPJ).

In late 2006 and early 2007, the CPJ and CFP published columns from the same authors that described a case of infant mortality caused by opioid overdose from breastfeeding attributed to the mother’s status as a rapid metabolizer of codeine to the active morphine metabolite.^1,2 The original version of this case report was published in 2006 in the Lancet.³ This case report has been cited over 600 times since its publication and its findings have had a significant effect on the way that postpartum analgesic medication is prescribed.⁴

In May 2020, Drs Robert Zipursky and David Juurlink from the University of Toronto published a paper in the journal Clinical Pharmacology and Therapeutics calling into question that newborns can develop opioid toxicity from breastfeeding.⁵ In the paper they reexamined aspects of this case report and strongly argued why such an occurrence is highly implausible.⁵ They concluded that

this explanation is implausible based upon several factors: (1) the exceedingly small amount of opioids passed into breastmilk irrespective of maternal CYP genotype, (2) the observation that significant neonatal opioid accumulation can only occur in the setting of severely impaired renal function, and (3) the previously unreported finding of a markedly elevated codeine concentration in postmortem blood. Finally, a review of the literature identifies a paucity of convincing reports of neonatal opioid toxicity during breastfeeding, with no other confirmed cases of neonatal death despite the use of these drugs by millions of nursing mothers over the past 2 decades.⁵

Around the time of the publication of this paper, Dr Juurlink contacted us asking that our respective journals retract the columns based on the new evidence that he and Dr Zipursky had brought to light. We reviewed the original Lancet case report, the columns, and Drs Zipursky and Juurlink’s article and together agreed that there were sufficient concerns about the validity of the case report findings that we invited Dr Koren and his colleagues to respond to the concerns raised by Drs Zipursky and Juurlink, which he did.

Given the complexity of the science, the significance of the case report, and the serious implications of retraction, we sought an independent peer review of the columns and case report; the Zipursky and Juurlink paper; correspondence from the parent of the child; and the responses of Dr Koren and colleagues to the concerns about the case report raised by Drs Zipursky and Juurlink’s paper.

We asked 2 international peer reviewers with expertise in pharmacology, pharmacokinetics, and pharmacogenomics to advise us if the case reports met the COPE (Committee on Publication Ethics) criteria for retraction.⁶ We report here that the reviewers independently reached the conclusion that the CPJ and CFP papers met these criteria. Both reviewers concluded that these columns met the COPE retraction criterion of “clear evidence that the findings are unreliable, either as a result of major error (eg, miscalculation or experimental error).”⁶

One of the reviewers concluded that there were additional grounds for retraction based on ethical concerns that the patient and family could be identified in the report (ie, that insufficient steps were taken to protect their identity).

We agree with the findings of the reviewers and therefore retract the papers:

• Madadi P, Koren G, Cairns J, Chitayat D, Gaedigk A, Leeder JS, et al. Maternal codeine is not safe for all breast-fed infants. Can Pharm J (Ott) 2006;139(6):56-60.

• Madadi P, Koren G, Cairns J, Chitayat D, Gaedigk A, Leeder JS, et al. Safety of codeine during breastfeeding. Fatal morphine poisoning in the breastfed neonate of a mother prescribed codeine. Can Fam Physician 2007;53:33-5.

• Copyright© the College of Family Physicians of Canada