This story is part of CBC Health’s Second Opinion, a weekly analysis of health and medical science news emailed to subscribers on Saturday mornings. If you haven’t subscribed yet, you can do that by clicking here.
A controversial Canadian medical paper, which pinned a baby’s death on codeine passed through breastmilk, is under renewed scrutiny a full two decades after it was published. The paper sparked a sweeping shift in global approaches to pain management and breastfeeding guidance for new mothers.
The Lancet, a leading medical journal, has now added an “expression of concern” to the 2006 case report after “new allegations of falsification of toxicological data, authorship issues, and ethical concerns” were flagged to the journal on Jan. 20. The move follows the recent publication of a year-long New Yorker investigation into the highly criticized paper, on top of years of Canadian media coverage.
Though outside researchers say the paper has long been debunked — and two other medical journals have already retracted similar versions — the case study has already been incredibly influential, leading to government warnings, changes in medication labelling, shifts toward the use of more potent and addictive forms of opioids, and untold numbers of women being told to choose between a common form of postpartum pain relief and safely breastfeeding their newborns.
The Lancet case study focused on the poisoning of an infant in Ontario in 2005, and purported that the baby boy’s mother had been prescribed Tylenol 3 and passed a deadly amount of morphine to her son through her breast milk.
The combination drug, commonly given for postpartum pain management, contains both acetaminophen and codeine, a mild opioid that gets partially metabolized into morphine inside the body. (Some individuals, including the woman in the case study, are genetically predisposed to converting codeine faster, and in larger amounts.)
For years after its publication, author Gideon Koren — the once-revered founder of the shuttered Motherisk drug testing lab — insisted the case showed maternally ingested codeine can be deadly for breastfeeding infants, despite mounting concern over Koren’s interpretation and questions from other scientists about its plausibility.
Tainted Tests: Broken Families
Koren has long been in the spotlight amid a slate of allegations of flawed or falsified study findings and a series of scandals surrounding the Motherisk lab.
The former facility at the Hospital for Sick Children (SickKids) in Toronto initially gained worldwide attention for using hair strands to test for drug and alcohol use in mothers, but its testing methods were deemed “inadequate and unreliable” by an independent provincial review in 2015.
Even so, the lab’s discredited tests were used in at least eight criminal cases and thousands of child protection cases — in many instances leading to the removal of children from their families — as the Toronto Star reported in 2017.
The Motherisk lab was shut down in 2019, and Koren agreed to relinquish his medical licence in Ontario that same year. He has not yet responded to CBC News’ request for comment.
Many long-time critics maintain that one of his most influential findings, from the 2006 case study in the Lancet, is flat-out impossible and should be fully retracted.
“In [Koren’s] telling, it was a fluke of genetics. Mum was taking codeine. She was turning it into morphine very efficiently. It was going to breast milk, and that’s how the baby died,” recalled David Juurlink, a well-known Canadian pharmacologist, toxicologist, and drug safety researcher. “And I believed that narrative for a few years.”
But it wasn’t long before Juurlink and others soon became unsettled by the study’s stunning result. Since then he’s raised alarms, saying the report falls apart under scrutiny.
Koren misinterpreted the toxicology results, Juurlink said, and didn’t appear to recognize that the extremely high acetaminophen and codeine concentrations in the infant’s bloodwork are “impossible consequences” of breastfeeding. New Yorker journalist Ben Taub noted that a postmortem examination showed the baby’s stomach contained “white curdled material,” with forensic testing finding codeine but not morphine — a result that, along with the “sheer magnitude” of drugs in his blood, suggested direct administration.
Yet the claims that codeine-laced breast milk can cause a baby’s death have reverberated throughout global public health messaging ever since Koren’s paper was published.
In 2008, Canada’s top pediatricians and obstetricians began developing new guidelines on which painkillers are safe during breastfeeding, while Health Canada issued warnings for nursing mothers and urged drugmakers to update codeine-based medication labeling to highlight the supposed risks. (Neither Health Canada nor the Society of Obstetricians and Gynaecologists of Canada provided a response to CBC News by deadline.)
Juurlink first called for a Lancet retraction in 2020. It asked for an investigation by SickKids, which concluded at the time that there was “no breach of lack of rigour, and that there is no justification for a retraction for an ongoing scientific dispute,” the journal noted in its new expression of concern.
Yet in 2020, two Canadian medical journals retracted similar papers Koren had published based on the case report, following Juurlink’s concerns.
The original case had been cited more than 600 times since its publication and its findings had a “significant effect on the way that postpartum analgesic medication is prescribed,” their joint retraction note said. Outside peer reviewers recruited by the Canadian journals ended up providing clear evidence that the findings were unreliable.
The Lancet said it has referred the latest allegations about Koren’s paper to SickKids’ research integrity office for a new investigation. A spokesperson for the hospital told CBC News that SickKids will be meeting with the journal and, “if warranted based on that discussion,” will initiate a formal review.
Physician and researcher Dr. Nav Persaud, an assistant professor at the University of Toronto and a longtime critic of other areas of Koren’s research, questioned why it’s taking so long for the influential medical journal to act on years of mounting concern about his work.
“I still don’t understand why the Lancet is saying that they need the institution — in this case, SickKids Hospital — to investigate before they would issue a retraction,” Persaud said.
This issue needs to finally be dealt with “in a definitive way.”
“This is a really tragic case, and you don’t want to lose sight of the fact that this particular issue started when a newborn tragically died, and the investigation of that newborn’s death has been botched.”
As the case study makes headlines yet again, Juurlink stressed that Koren’s influence has led to a generation of physicians and parents believing that babies can be harmed, or even die, due to a breastfeeding mother taking prescribed opioids to ease the pain of a C-section or episiotomy. He worries that’s led to many mothers avoiding breastfeeding entirely, despite the benefits.
“One of the more disturbing aspects of this whole saga is it’s very easy to find instances of infants who have died, obviously from homicide — given a drug, given an opioid, by a caregiver — and experts in the court system blamed breast milk,” he said.
“It’s basically a myth, and it all goes back to one sloppy misreading of a single case 20 years ago.”
