STATS ARTICLES 2008
Prozac Wars: Slate versus “The Infinite Mind”
Trevor Butterworth, May 12, 2008
A vicious spat over sourcing and disclosure on an NPR- aired show reveals two news organizations at fault.
In the fall of 2004, the Food and Drug Administration (FDA) mandated a “black box” warning on Prozac and similar drugs because emerging research suggested that there might be a link between taking such antidepressants and increased suicidal thoughts and behavior in teens. Soon, however, research showed that there was a spike in the number of adolescent suicides after doctors cut back on prescribing these drugs, known as SSRIs (serotonin selective reuptake inhibitors).
The new data on suicide and prescription patterns, which was published in the September, 2007 issue of the American Journal of Psychiatry, raised an difficult question: did the FDA warning, given widespread coverage by the media, lead to more suicides than it was intended to prevent?
This question is now the backdrop to a debate now raging between the online magazine Slate and "The Inifinte Mind," a radio show aired on some NPR stations. The Infinite Mind broadcast a program in March looking at whether excessive panic after the black box announcement, generated in part by media reporting, was responsible for a false sense of risk from SSRIs. And then in an article last week for Slate, journalists Shannon Brownlee and Jeanne Lenzer argued that the program was biased because all its experts, and even the show's host, had at some point received money from the pharmaceutical industry, none of which was disclosed.
It is difficult to underestimate the importance of the issues raised by this debate, given that precautionary thinking in regulation is being given more and more credence in America, especially at the political level. The precautionary principle is both seductive and saleable to the public: if there appears to be a risk, then caution seems a highly rational direction to go in, legislatively speaking; and one that happily requires little in the way of a complex risk assessment and justification to a non-scientific audience.
But to bank everything on the idea that you shouldn’t do something because it carries a risk without considering that there may also risks to not acting is, in the end, not very sensible. And this is particularly true of medicine where most treatments for complex problems carry some, and sometimes substantial, risks.
Defending SSRIs on the Infinite Mind
In trying to redress the imbalance on SSRIs created by the black box warning and the subsequent media coverage, the Infinite Mind, unfortunately, appears to have over-compensated in the other direction: In introducing its program, “Prozac Nation: Revisited,” host Fred Goodwin categorically stated, “There is no credible scientific evidence linking antidepressants to violence or to suicide.”
This statement doesn’t accurately reflect the complexity of the research on this subject. There are thousands of reports of adverse reactions to these medications-- including a side effect called “akasthisia,” in which the patient becomes extremely energized, agitated, impulsive, unable to sleep and may report desire to harm himself or others.
In clinical trials of paroxetine, for example, about .5% of patients experienced a “hostile event,” in which they either had thoughts of harming someone else or behaved aggressively. And research giving these medications to people without depression finds that it does produce suicidal ideation in some cases.
Because both homicide and suicide are (thankfully!) rare, it is impossible to do clinical trials large enough to determine whether these are more common amongst people taking SSRI medications. However, there have been enough incidents in which people starting these medications, changing dosage or stopping them reported a desire to hurt themselves or others that they had not experienced previously, that the FDA required a “black box” warning on the drugs related to suicidal thoughts in adolescents.
The “Infinite Mind” guests linked this FDA decision to hysteria –and correctly noted that after the “black box” was mandated, prescribing rates for teens went down and teen suicide went up. The host and guests noted research finding that suicidal thoughts aren’t correlated with suicide itself (largely, it is believed, because the most suicidal people are intent on doing it so they don’t reveal their thoughts to others who could stop them)-- suggesting that the medications increase only suicidal thoughts, not actions.
However, while it is true that for most people, these medications are more likely to prevent suicide or homicide than to drive them to it, it also appears to be the case that in a small minority, starting, stopping or changing dosages of these meds can indeed produce suicidal or homicidal thoughts and even behavior.
The show implied that science had proved that the homicides committed by Stephen Kazmierczak at Northern Illinois University weren’t caused by his cessation of antidepressant medication because in most cases, the medication reduces risk of suicide or violence. But this is not the case: It is possible that he was indeed one of the rare cases in which stopping antidepressants produced homicidal acts. Of course, whether that was due to drug withdrawal or to a return of the original depression is impossible to tell – but in its efforts to show the overall good done by these medications, the show failed to adequately discuss the rare cases where they can do harm.
The data does not show categorically that these drugs can’t occasionally “backfire” and it is a disservice to listeners to claim otherwise in an attempt to “balance” media coverage that goes too far in emphasizing the negative. Including an interview with researcher David Healy - who is widely known for his work on antidepressant side effects but does not oppose all use of the drugs - or with someone else who has studied problems associated with these drugs would have not only added balance to the show, it would have made it much more interesting and informative.
Attacking Pharma on Slate
Unfortunately, Slate magazine’s “Medical Examinerr” did little in the way of examining the medicine, when it launched an attack on the show’s credibility by pointing out that all the sources involved in the discussion received money from the pharmaceutical industry. In “Stealth Marketers,” Shannon Brownlee and Jeanne Lenzer noted:
“The radio show, which was broadcast nationwide and paid for in part by the John D. and Catherine T. MacArthur Foundation, had the air of quiet, authoritative credibility. Host Dr. Fred Goodwin, a former director of the National Institute of Mental Health, interviewed three prominent guests, and any radio producer would be hard-pressed to find a more seemingly credible quartet. Credible, that is, except for a crucial detail that was never revealed to listeners: All four of the experts on the show, including Goodwin, have financial ties to the makers of antidepressants. Also unmentioned were the "unrestricted grants" that The Infinite Mind has received from drug makers, including Eli Lilly, the manufacturer of the antidepressant Prozac."
And that was, more or less it; in effect, the Infinite Mind is one of the worst examples of a worsening trend in media reporting on health issues, the failure to disclose industry ties. Nowhere in the piece do the writers address the need for antidepressants, or the fact that reduced prescribing in the wake of the black box warning pushed up the teen suicide rate. This, in its own way, is just as bad as what they accuse the Infinite Mind of doing: it breaks faith with science.
Lenzer, a former physician’s assistant turned freelance writer, appears to be of the “industry money corrupts” school of medical reporting. This has its place, but only if an accusation of vested interests can be shown in faulty research methodology and poor statistical methods, or that the drug is pointless or has a greater risk profile than similar drugs without any greater benefit. Otherwise, it’s a way of reporting whose bias is at least as bad as the bias it purports to expose precisely because it’s unscientific in method; it insinuates corruption without demonstrating it in the data.
The reality is that the high cost of doing rigorous research in medicine is largely supported by industry, and that, paradoxically, “independent” research can sometimes be less reliable simply because there isn’t the funding to generate large sample or multi-generational studies. Peer-review, international protocols, independent auditing of laboratories, objective statistical criteria, and the crippling penalties for getting it wrong do much to make industry-funded research more honest than popular culture, fed by movies like the “Constant Gardener,” is willing to believe it is.
So, of course, the Infinite Mind should have been up front with the financial links of the guests and the program. But the reality of scientific research means that the “follow the money” habit of the press demands a concomitant obligation to follow the statistics. The problem is that this puts mere journalists in the position of sitting in judgment on data they have they would seem to have little qualification to judge.
Slate’s article generated a stiff response from the Infinite Mind’s Senior Executive Producer Bill Lichtenstein (which can be read following the Slate article). He reminded Slate readers that all but one of his guests ranked as a leading expert in the field, and ended with a kicker that propelled the dispute out of science and onto Jim Romenesko’s media column at the Poynter Institute.
“In the interest of full disclosure, I also should note for the record that Lenzer, who co-authored the Slate article, called me a few days after the "Prozac Nation: Revisited" program aired to pitch a program that she wanted us to do for The Infinite Mind, called "Journalists on Prozac," which would feature her and her writing partner Shannon Brownlee. Checking into Lenzer's credentials, I found a troubling article in The New York Times taking her to task for a British Medical Journal article that suggested that Eli Lilly and Company, which makes Prozac, had concealed documents about the link between anti-depressants, suicide and violence. The BMJ subsequently retracted the article, with full apologies, and the whole matter was widely covered in the news media.
After we told Jeanne Lenzer that we would not be proceeding with a program featuring her, she and Brownlee wrote the article for Slate.”
This, naturally, triggered a response from Brownlee and Lenzer, who again argued that both the failure to disclose financial support and the support itself corrupted medical research; they also disputed the idea that they’d pitched themselves to the Infinite Mind. Again, both writers failed to note any science to suggest that antidepressants posed a risk or to contradict the association between the black box warning and the increased incidence of suicide. Lenzer also dismissed the accusation about her BMJ article:
“Finally, the BMJ article to which Mr. Lichtenstein refers involved a correction of a minor point – not a substantive one. In fact, the full truth about that story is one we plan to tell in another forum and a link HERE will be inserted when that article is published.
As for Lenzer's credibility as a journalist, readers might like to know that the BMJ has continued to publish her articles on a regular basis. So have other respected outlets.”
But Lenzer’s piece in the BMJ resulted in a bit more than a minor correction. Lenzer claimed she had received documents from an anonymous source that Eli Lilly had failed to disclose in a product liability trial over Prozac, and to the FDA, and which purported to show a link between the drug and suicidal ideation and violent behavior.
Not surprisingly, this “minor point,” appeared to be more of a smoking gun – at least to other journalists, who jumped all over the story of missing documents showing that Prozac was more dangerous than had been thought. But in her zeal to prosecute a company for apparently burying data harmful to its financial health, Lenzer took a gamble: she did not show the documents to Eli Lilly. As it turned out, they had been circulating for years and had been disclosed at the trial and to the FDA. The British Medical Journal conducted an investigation, retracted the claim and apologized to the company. As the claim was, in effec,t the story, the retraction effectively meant that the entire story was retracted.
This all happened three years ago, and one might wonder why it has taken so long to tell the “full truth” or what the “full truth” could in fact be. But a “correction of a minor point” this is not. Journalists have lost gigs for much less. Perhaps Lenzer was saved by virtue of the fact that she withheld the documents with the apparent agreement of her editors at the BMJ.
To sum up: one can be too credulous of industry or dismissive of its importance; but one can also be too cynical and lurch towards paranoia when investigating its influence. Neither disposition helps the public understand the science. In the case of SSRIs, the Infinite Mind needs to deal with the fact that it underplayed the risk; but Slate and Brownlee and Lenzer need to consider something less palatable: whether their approach to reporting this issue is putting, on balance, more lives at stake.