On April 29th, 2005, a newborn and formerly healthy baby boy named Tariq Jamieson suddenly stopped eating. As the day progressed, his parents became increasingly concerned. A child that age should be eating every few hours. By evening, scared that something was wrong, his parents called a telehealth service. An ambulance was sent to their home. Shortly before it arrived, Tariq stopped breathing. He was pronounced dead that night at a top hospital in Toronto, Canada.
An unspeakable tragedy had just befallen this family.
Why would a healthy newborn die in such a strange manner? He had been eating and sleeping fine. For a child to simply stop eating, sleep for most of a day, and then die is almost unheard of. Given the unique circumstances, a full autopsy was performed. The results were surprising. Tariq had a lethal amount of codeine and morphine in his bloodstream.
A specialist and world-renowned expert, Dr. Gideon Koren, was brought in to determine what happened. He quickly made the connection that Tariq’s mother, Rani, had been prescribed Tylenol with codeine after birth. At the time, this wasn’t uncommon. Women in North America were often prescribed codeine or other opiates after a difficult delivery. It is known that codeine metabolizes into morphine in the liver, and that some of that morphine can pass into breast milk and then into a newborn. The dose, however, would typically be far too low to cause any reaction, let alone harm or death, in a newborn.
Some women, however, have a genetic mutation that causes their liver to metabolize codeine into morphine at a much higher rate. Rani Jamieson took a DNA test and had that mutation. Some of Rani’s breast milk was also tested and found to have a morphine concentration of 87 nanograms per milliliter, far higher than typical. Dr. Koren quickly published a paper explaining the case study in The Lancet, a top medical journal.
The implications were massive. Millions of women were being prescribed codeine or other opioids after giving birth. Around five percent of American women have the genetic mutation that supercharges codeine metabolization. If there was a chance, even a small one, that the codeine could be converted into morphine at a high enough dose to kill an infant, prescription guidelines needed to be changed immediately. Dr. Koren quickly spread the alarm, becoming an evangelist against codeine. Health organizations adopted new standards. In 2017, the FDA restricted the prescription of codeine in children and strongly recommended against it in breastfeeding women. Change rippled around the world, and codeine went from a commonly prescribed post-partum medication to a rarity.
It was all for naught.
As expertly relayed by Ben Taub in a recent New Yorker article, the findings of Dr. Koren made no sense. Yes, Rani Jamieson had the genetic mutation that could have resulted in a high amount of morphine in her breast milk. Yes, Tariq died of a morphine overdose. But the amount transferred simply wasn’t possible. Rani had a morphine concentration of 87 nanograms per milliliter in her breast milk. But in another case, a mother was found to have 500 nanograms per milliliter in her breast milk. Even in that extreme instance, a typical feeding would result in a roughly 0.05 milligram morphine dose to the child. And morphine, while potentially dangerous, is still prescribed to babies in pain. A four-kilogram newborn in a hospital could be prescribed 1.00 milligrams of morphine orally. In other words, the amount of morphine that could have been transferred to Tariq from Rani via breast milk was more than an order of magnitude less than a fatal dose. Tariq’s concentration was more than fifty times the expectation. It’s like an otherwise healthy adult getting fatal alcohol poisoning from drinking a single beer. Something else was going on.
It took a dinner between Dr. Nick Bateman, then the director of the Scottish branch of the U.K.’s National Poisons Information Service, and Dr. David Juurlink, a toxicologist, to get the ball rolling. Bateman had been stewing over The Lancet article for some time. There was just no way for a baby to have such a high concentration of morphine in their system from codeine taken by their mother, regardless of genetics. He eventually dropped a bombshell: “David, whatever the intention, that baby was poisoned.”
To his credit, Dr. Juurlink took that ball and ran with it. He spent years developing an argument and developing evidence that Dr. Koren’s conclusions were wrong. This takes serious fortitude. Dr. Koren was a world-class expert. Even in the best of circumstances, challenging such a titan in the sciences is risky. Also, Dr. Juurlink is critical of opioid use and has said medical professionals must accept blame for fentanyl overdoses. Yet here he was, taking on a superstar in his field whose broad conclusions agreed with his own. But the truth matters. If newborns can be fatally poisoned by morphine that was metabolized in the mother, that matters. It’s yet another reason to limit opioid use. But if newborns cannot be fatally poisoned in that manner, that matters too. New mothers should have their pain properly treated, and opioids, for all their flaws, are fantastic at treating pain.
Dr. Juurlink quickly realized nothing about the Tariq Jamieson case added up. First was the morphine disparity, which, again, was impossible to reconcile with the facts. Then, additional evidence came to light. Not only was Tariq found to have an elevated amount of morphine and codeine in his system, but he was also found to have an elevated level of acetaminophen, the main component of Tylenol. This made even less sense than the morphine and codeine. Acetaminophen can’t easily pass from the digestive tract of a nursing mother to her breast milk. Tariq’s concentration of acetaminophen was around two hundred times the expected level. On top of that, the coroner’s case files showed that Tariq had two milliliters of “white curdled material” in his stomach, and that substance tested positive for codeine. For a newborn who hasn’t eaten in 18 hours, no milk can remain in the stomach for nearly this long. Dr. Juurlink came to a clear conclusion: Tariq had been fed Tylenol with codeine. Only that could explain the concentration of morphine in his blood.
In a bizarre twist, Dr. Juurlink also spoke with Dr. Michael Rieder, who co-authored a separate article with Dr. Koren about a second child, nicknamed “Baby Boy Blue”, who overdosed on morphine via breast milk, but survived after doctors in the hospital administered naloxone. This indicated that maybe Dr. Juurlink was missing something in the Tariq Jamieson case. Two babies suffering from the same rare condition aren’t a lot, but it could provide missing information that explains how Tariq’s morphine concentration was so high. To Dr. Juurlink’s astonishment, when asked, Dr. Rieder said, “Oh, we made it up.”
The entire case was fictional, but there was no indication of this in the published article. Anyone reading it, including other medical professionals, would assume these were true incidents. After an outcry, the journal issued corrections on all 138 case reports in the series the “Baby Boy Blue” article was published in, describing them as fictional. This came as a surprise to Dr. Farah Abdulsatar, who published one of the other 137 case studies in the series. Hers, however, was true. There was apparently a gentlemen’s agreement that the cases could be fictional, but this was only known to those with inside knowledge. To this day, it isn’t clear how many of these case studies, all published in a legitimate medical journal, are fictional.
Dr. Juurlink published his own study in 2020, titled “The Implausibility of Neonatal Opioid Toxicity from Breastfeeding.” The conclusion is right there in the title. Newborns cannot consume a fatal dosage of morphine from breastfeeding.
The circumstances surrounding Dr. Koren, the doctor who originally published the article about Tariq Jamieson, went from strange to outright bizarre. In the 1990s, long before the Tariq Jamieson overdose, he sent a series of anonymous deranged letters to colleagues who disagreed with him about a drug’s efficacy. Those who knew Dr. Koren believed it had to have been him who sent the letters, but he denied responsibility. It was only after DNA evidence from licking the envelopes proved his guilt that he admitted sending them. In 2015, Dr. Koren’s lab, Motherisk, was shut down after it was found to have performed thousands of tests without following standard procedure. At least one of these faulty tests led to a mother being falsely imprisoned. In 2019, Dr. Koren surrendered his medical license and agreed to never practice medicine in Ontario.
He was also fiercely protective of his publications, at one point trying to block an article calling his Lancet publication into question by claiming the rebuttal was defamatory and the product of someone being paid by Johnson & Johnson, the maker of Tylenol. “It is one thing to philosophize in a journal, it is another thing to knowingly injure a grieving family,” Dr. Koren wrote to the editor at The Lancet. In other words - never mind science, won’t someone think of the children? He also authored an implausible two thousand academic articles, at one point publishing more than a paper a week. Many of these have since been retracted.
Yet despite all the scandal, the recommendations remain. The FDA still recommends against prescribing new codeine to nursing mothers. Other organizations around the world do the same. This despite the retraction of numerous articles, aberrant behavior, and the medical banishment of Dr. Koren. That’s bad enough, but the underlying recommendations themselves are mystifying.
To those in the medical community, what the hell are you doing?
Even if Tariq Jamieson died from morphine in breast milk. Even if the paper Dr. Koren wrote was totally legitimate, why on earth would that result in any significant recommendations? It’s a single case study! Even a significant research effort that surveys hundreds of patients shouldn’t have this kind of impact. I expect individual researchers to make mistakes like this. But not the entire medical community. Doctors rarely commit fraud à la the Andrew Wakefield autism paper, but they make mistakes all the time. Blood tests are incorrect. Patients misremember. Researchers flip signs and make math mistakes. These things happen. All. The. Time. In every field.
That’s why science builds on a body of work. Why dozens of researchers ask the same question from different angles and in different circumstances. Over time, a body of literature develops that generally, but not always, points towards the truth. It is simply unthinkable that one case study could result in entirely new guidance on the use of a commonly prescribed medication. Common sense shows that morphine overdose via breast milk doesn’t happen this way. Five percent of American women have the genetic mutation that increases morphine metabolism. If this was dangerous, there would be hundreds of newborns overdosing on morphine every year. The risk would have been identified decades ago.
It is defensible for a single bad paper to avoid detection and even become well-cited. Things happen. People defer to the experts, and most don’t have time to check the math of their colleagues. It is not defensible for the entire medical industry to develop guidelines based on a single case study.
I honestly cannot fathom how these decisions were made. How do the FDA and other organizations decide that medical guidance should rest on a single instance of an adverse reaction? People die every year from having wisdom teeth removed and taking birth control. That doesn’t mean those things are banned. Nor should it. It’s the responsibility of medical organizations to make informed decisions. To take into account costs and benefits. To recommend against an important pain medication because of a single reported death is obscene. It goes against science. It goes against common sense. And it makes one question how other decisions are being made.