Review of Children of the Cure, written by Healy, Le Noury, and Wood
Review by David Antonuccio, Ph.D.
Professor Emeritus, Dept. of Psychiatry and Behavioral Sciences,
University of Nevada, Reno, School of Medicine
I suspect the title of this book by David Healy, Joanna Le Noury, and Julie Wood, was inspired by the Stephen King horror story Children of the Corn. In the Stephen King story, it is the children who are evil. In Dr. Healy’s book, it is the adults who are perpetrating evil by exaggerating benefits and minimizing harm of antidepressant use in children. Specifically, this book tells the story of Study 329, a study that is published in the top adolescent psychiatry journal and concludes that Paxil, manufactured by GlaxoSmithKline (GSK), is both safe and effective for adolescent depression (Keller et al., 2001). As it turns out, these conclusions are false. After a reanalysis of the raw data under the RIAT (Restoring Invisible and Abandoned Trials) initiative, the actual results of Study 329 show that Paxil is not more effective than placebo and causes more harm than placebo in depressed adolescents who are randomly assigned to treatment (LeNoury et al., 2016). Really!
Since the publication of the reanalysis of Study 329, it appears this is the first study in the world to be published in two different prestigious peer reviewed journals with exactly opposite conclusions. The same study! How did this happen?
The truth is that the study was “ghost designed”, “ghost analyzed”, and “ghost written”. This truly is a “ghost” story that harms children and should scare parents.
As in most antidepressant trials, placebo responders were kicked out of the study before it began. Also, the original study was largely ghost written by previously unidentified GSK officials (named in the book) and the ostensible named authors on the original article may not have actually had access to the raw data as it was sequestered by the manufacturer. Poor outcomes on the primary measures were ignored and secondary outcomes were highlighted. Among other coding tricks, suicidal ideation or behavior was euphemistically coded as “emotional lability” and violent ideation or behavior was coded as “hostility”, obscuring serious adverse events. Honestly, the term “ghost writing” does not do justice to what actually happened. The truth is that the study was “ghost designed”, “ghost analyzed”, and “ghost written”. This truly is a “ghost” story that harms children and should scare parents.
Unfortunately, I know from first-hand experience how this can happen because I was once a coinvestigator in a multi-site study funded by Marion Merrill Dow on the safety of nicotine replacement treatment (NRT) for cardiac patients who were trying to quit smoking (see https://usatoday30.usatoday.com/news/health/2001-05-17-drug-companies.htm). Our study was essential to the manufacturer’s goal of making the nicotine patch available over the counter. The scientists on our team insisted on evaluating outcome as well as safety. About 600 patients were randomly assigned to nicotine patch or placebo patch for 10 weeks, with minimal (about a total of 15 minutes) behavioral counseling. Over my loud protests, someone (the GSK company handlers?) insisted on separating the study into a safety paper and an efficacy paper. The safety paper, which found no increased no increased health risk for the patch group, was published in the New England Journal of Medicine (Joseph et al., 1996). The efficacy paper was never published. This was unfortunate because at 48 week follow-up the placebo group had a nonsignificantly higher quit rate (12%) than the patch group (10%). The point is that our study showed that the patch was safe but not effective in this group of cardiac patients. Our study confirmed what I had suspected, that the patch with minimal behavioral skills training did not help patients quit smoking any better than a placebo. I felt our published paper in a top journal was potentially misleading because it implied that the patch alone might be effective.
I spent the next 3 years trying to get independent access to the data so I could analyze and publish the outcome data myself. I requested the data from the company research coordinator, complained to the lead investigator, notified the Veterans Administration research office, and threatened a FOIA action. Even though I was an identified coinvestigator, nothing worked. Until I wrote to the editor of the NEJM explaining what had happened and explaining that their readers had a right to know the outcome results of our study. The NEJM editor offered to publish my letter if I were not given access to the raw data. Shortly thereafter, the lead investigator sent me the raw data and we were able to collaborate on a brief report in the NEJM highlighting the lack of efficacy (Joseph & Antonuccio, 1999). I then dropped my complaint. But the damage had been done. By then, NRT was widely available over the counter and consumers could directly access this medication without behavioral skill training despite the negative outcome of our study. To be sure, any effort to quit smoking was a good thing, but volunteer human subjects make the sacrifices they do so that all humans can have access to and be informed by the data they generate (see Antonuccio & Healy, 2008). Our study fell far short of that goal. Though it would be difficult to prove after all these years, I now believe, in hindsight, that the initial draft of our study was ghost written and the raw data were ghost analyzed, just like Study 329.
Flash forward to 2004. I along with psychologist Irving Kirsch and then graduate student Amanda Drews, analyzed the entire literature of published controlled antidepressant trials in children available at the time, an analysis that included Study 329. We presented our analysis at the FDA hearings on the safety and efficacy of antidepressants in children in Bethesda, Maryland on February 2, 2004 (AntiDepressants, FDA2-2-2004 Hearing ). Our analysis showed that placebo duplicated 87% of the active drug response, suggesting that differences between drug and placebo were not clinically meaningful. Subsequent analyses have backed up this conclusion (Antonuccio, 2008). Our position was that, if the medications were not more effective than placebo, no extra risk was warranted. A drug would have to at least offer some benefit to warrant even minimal risk. Antidepressants in children did not clear that low bar. As an ironic twist to these hearings, Healy, the world’s leading expert on suicidal risk associated with antidepressants, was not permitted to testify for reasons that are still a mystery to me. He surely could have shed more light on this scientific literature. In any case, the final outcome was the implementation later that year of FDA Black Box Warnings, warnings that organized psychiatry has been trying to undermine since that time (see Spielmans, Spence-Sing, & Perry, 2020), despite the fact the regulatory bodies of Canada, Europe, Australia, and the UK all independently came to the same conclusions about the increased risk of suicidal behavior caused by these medications.
As I began reading Healy’s book, I noticed that one of my old classmates was an author on Study 329. I did not know this classmate well but I respected his work and we had exchanged friendly emails in the past. So I contacted him by email to see if he might be willing to chat with me about Study 329. I can’t say for sure whether he received my email, but I never heard back from him. After reading Children of the Cure, I now understand why he might have been reluctant to respond. It appears the original authors of Study 329 were either duped or complicit in a publication that exaggerated the benefits and minimized the harms of Paxil in adolescent depression. The U.S. Justice Department fined GSK $3 billion dollars for this study and other deceptive research practices. Children who were exposed to this drug, paid an unknown price in side effects and harm. The authors of the original publication will likely escape any meaningful consequences. I want to be charitable and hope that the authors, or at least my former classmate, were duped as I was in the one drug company trial in which I participated. But to come to that conclusion would require the authors to stop defending their bogus conclusions in the deceptive and disingenuous manner detailed in the Healy book. That would at least be a start.
The Healy book dives deep and uncovers the truth about Study 329.
The Healy book dives deep and uncovers the truth about Study 329. But patients are now left with conflicting conclusions from the same study and the guidance from doctors who may not fully understand the implications of the reanalysis of Study 329. The truth is that the most vulnerable patients are the most compliant, those who feel unable or unwilling to question the authority of the doctor who recommends a medication like Paxil that has now been discredited by the restored analysis. If the patient does have the audacity to raise questions or report serious side effects, they may encounter anger or resentment from the doctor who is caring for them. Quoting from the last chapter of the Healy book, “Today’s doctors are better trained to break bad news, but not to hear bad news. Underneath the surface of many otherwise decent people in healthcare lies someone who thinks they are an expert and who, convinced of their own good intentions, figures the primary duty of the person who comes to them is to do as they are told. But how can anyone be an expert given the kind of knowledge that tumbles out of studies like Study 329?” I heartily agree!
Review by David Antonuccio, Ph.D.
Professor Emeritus, Dept. of Psychiatry and Behavioral Sciences,
University of Nevada, Reno, School of Medicine
References
- Antonuccio, D. (2008). Treating depressed children with antidepressants: more harm than benefit?. Journal of Clinical Psychology in Medical Settings, 15(2), 92-97.
- Antonuccio, D. O., & Healy, D. (2008). The researcher’s credo. British Medical Journal, 336(7645), 629-629.
- Joseph, A. M., & Antonuccio, D. O. (1999). Lack of efficacy of transdermal nicotine in smoking cessation. New England Journal of Medicine, 341(15), 1157-1158.
- Joseph, A.M., Norman, S., Ferry, L., Prochazka, A., Westman, E., Steele, B., Sherman, S., Cleveland, M., Antonuccio, D.O., Hartman, N., McGovern, P. (1996). The safety of transdermal nicotine therapy as an aid to smoking cessation in patients with cardiac disease. New England Journal of Medicine, 335(24), 1792-1798.
- Keller, M.B., Ryan, N.D., Strober, M., Klein, R.G., Kutcher, S.P., Birmaher, B., Hagino, O.R., Koplewicz, H., Carlson, G.A., Clarke, G.N. and Emslie, G.J., 2001. Efficacy of paroxetine in the treatment of adolescent major depression: a randomized, controlled trial. Journal of the American Academy of Child & Adolescent Psychiatry, 40(7), pp.762-772.
- Le Noury, J., Nardo, J. M., Healy, D., Jureidini, J., Raven, M., Tufanaru, C., & Abi-Jaoude, E. (2015). Restoring Study 329: efficacy and harms of paroxetine and imipramine in treatment of major depression in adolescence. British Medical Journal, 351, h4320.
- Spielmans, G. I., Spence-Sing, T., & Parry, P. (2020). Duty to warn: antidepressant black box suicidality warning is empirically justified. Frontiers in psychiatry, 11,
Four Ways to Buy the Book
Buy the Book in Print:
- Amazon.com: Children of the Cure (Print)
- Lulu.com: Children of the Cure (Print)
Buy Book in Electronic Formats:
- Kindle: Children of the Cure (Kindle)
- PDF: Children of the Cure (PDF)
Leave a Reply