JAMA's landmark antidepressant study still doesn’t tell the whole story

Twenty years ago, a landmark JAMA study changed the way antidepressants were prescribed during pregnancy. Today, readers still aren't seeing the complete scientific record.

Twenty years ago, a paper published in JAMA changed the way doctors thought about antidepressant use during pregnancy.

The study concluded that women who discontinued the medication during pregnancy were five times more likely to relapse into major depression than those who remained on the medication. It quickly became one of the most influential papers in perinatal psychiatry.

doi:10.1001/jama.295.5.499

Its findings were incorporated into clinical guidelines, cited by regulators, and eventually referenced in FDA prescribing information.

For many clinicians, it became the definitive evidence that discontinuing SSRIs during pregnancy carried a substantial risk of depressive relapse.

At the time, maternal-foetal medicine specialist Dr Adam Urato noticed something was missing from the study’s disclosures.

Dr Adam Urato, Maternal-foetal medicine specialist

He recognised that many of the authors’ financial relationships with manufacturers of antidepressants had not been disclosed.

Urato wrote a letter to JAMA, published weeks later, pointing out that nine physician authors had financial relationships with antidepressant manufacturers, yet only two had disclosed them in the published paper.

“For readers to be able to make informed judgments about potential biases in this study,” he wrote, “they should have been made aware of all of these associations and potential conflicts of interest.”

The Wall Street Journal also published an investigation that documented more than sixty undisclosed financial relationships between the study’s authors and manufacturers of antidepressants.

The investigation revealed consulting fees, research funding, speakers’ bureau payments and advisory roles involving companies whose products were directly affected by the study’s conclusions.

The study’s authors initially defended the omissions. Because the paper was not evaluating any one manufacturer’s antidepressant, they concluded there was “no inherent conflict of interest” arising from their industry ties.

They nevertheless apologised for failing to disclose them.

“In retrospect,” they wrote, “we regret that we failed to include disclosures of the financial associations of all of the authors. Such disclosures would have provided utmost transparency.”

The controversy also caught the attention of the New York Times editorial board, which wrote a scathing editorial criticising the authors’ explanation for the omissions.

JAMA responded by publishing the expanded disclosures as a separate correction and later strengthened its conflict-of-interest disclosure policy.

The problem is that the correction was never incorporated into the original article. Twenty years later, anyone reading the paper online still sees the original, incomplete disclosures, with no indication that they were later corrected.

Instead, the updated disclosures remain buried in the journal’s correspondence pages.

When I asked JAMA why the online article had never been updated, the journal responded:

“This article was published 20 years ago, when we did not have the technical capabilities available today. At that time, the procedure was to append the correction notice to the end of the article PDF. That is what is displayed online.”

In a recent interview with MD Reports, Urato called JAMA’s explanation “ridiculous,” saying it failed to explain why the article remains unchanged today.

“We’ve been in an online era for more than 20 years,” he said. “Their argument that they didn’t have the technical capabilities is laughable.”

He says the issue centres on what readers see today.

“When you look this paper up online, you find the free full-text version of the original article, which is missing more than sixty financial disclosures,” he said.

“The original article is inaccurate, and there is no indication whatsoever that there is a correction to this paper.”

“The vast majority of readers will simply be misinformed.”

Urato says the solution is straightforward.

“JAMA needs to make it clear to readers, preferably with an obvious warning at the top of the article webpage, that there is a correction to this article, and JAMA needs to make that correction freely and easily available.”

More than missing disclosures

The financial disclosures are only part of the story.

The paper repeatedly attributes symptoms after stopping antidepressants to a relapse of depression. It never discusses the possibility that, in some women, those symptoms could instead represent antidepressant withdrawal.

“There’s no discussion about withdrawal at all in the paper,” Urato said. “They don’t mention the word withdrawal once.”

For Urato, that omission makes the undisclosed financial relationships all the more significant.

“The pharmaceutical industry wants those symptoms to be regarded as a relapse of depression,” he said. “That’s why readers need to know the authors were being paid by the manufacturers.”

The paper’s central finding has also proved difficult to replicate.

Subsequent studies have reported considerably lower and more variable relapse rates, and none has reproduced the dramatic five-fold increase reported in the original paper.

“Now, here we are twenty years later, after it’s influenced guidelines and been cited over and over again, and people accessing it online wouldn’t even know there was ever a correction,” he said.

Beyond one paper

For Urato, the JAMA paper is more than an isolated case. He believes the controversy reflects a wider problem throughout medicine.

“The organisations that are supposed to be working on behalf of the public are instead working for the pharmaceutical industry,” he said.

He cites organisations including the American College of Obstetricians and Gynecologists (ACOG), the American Academy of Pediatrics, regulators such as the FDA, and major academic institutions as examples of bodies that receive substantial industry funding.

“The public needs to understand this,” he said. “Everybody understands that money influences politics and business. But when it comes to medicine, people somehow think it doesn’t influence judgement.”

Twenty years after JAMA acknowledged the missing disclosures, the version of the paper that most readers see still does not tell the full story. For a study that shaped doctors' decisions about prescribing antidepressants during pregnancy, that is more than a historical oversight.

---