Editor’s Note: Over the next several months, Mad in America is publishing a serialized version of Peter Gøtzsche’s book, Critical Psychiatry Textbook. In this blog, he discusses how depression pills don’t work for children and increase their risk of suicide. Each Monday, a new section of the book is published, and all chapters are archived here.
Depression pills don’t work for children and double their risk of suicide
One of the textbooks mentioned a meta-analysis of 34 randomised trials of depression pills given to children and adolescents and claimed that fluoxetine was the only drug with a significant effect and also with the highest tolerability.19:215
Such claims belong to the section for science fiction. It is fairly impossible for a drug to be both more effective and less harmful than similar drugs from the same class. Even though this textbook had references, there were none to this implausible claim, but I believe the source can only be the 2016 network meta-analysis by Andrea Cipriani and colleagues.297
To increase the power of the analyses, the authors included both placebo-controlled trials and head-to-head comparisons, but they were not sufficiently careful. They mostly used published trial reports (only 7 of the included 34 trials were unpublished), which are substantially biased.2,7,8
As an example, statistician Hans Melander and colleagues, working for the Swedish drug agency, showed that placebo-controlled trials of SSRIs were more often published when the results were statistically significant, and that many publications ignored the results of intention-to-treat analyses and reported the more favourable per-protocol analyses where only patients who do not drop out of the study are retained in the analysis.314 This created a misconception about how effective the drugs are.6:137 Moreover, cross-references to multiple publications of the same trial were missing, and sometimes they had no author names in common and therefore looked like separate trials.
As another example, psychiatrist Erick Turner, who worked for the FDA, and colleagues noted that 31% of the trials done as part of a licensing application for SSRIs and related drugs viewed by FDA as negative or questionable were published as positive, and the effect size in the published articles was 32% higher than in FDA’s reviews of all the trials.315
Fluoxetine is unsafe and ineffective and the trials are manipulated
Fluoxetine was approved in the United States in 2002 for depression in children and adolescents based on two placebo-controlled trials, X065 and HCJE, with 96 and 219 participants, respectively, although FDA’s statistical reviewer had noted that there wasn’t a statistically significant benefit for the drug on the primary outcome in either trial.316 As both trials appeared to have been misreported in the literature, David Healy and I decided to restore them according to the RIAT initiative (Restoring Invisible and Abandoned Trials).317
We reviewed the 3557 pages of clinical study reports Eli Lilly had submitted to the regulators and found serious manipulations with the data, both in the reports and in the publications.279 Essential information was missing; there was contradictory information, even related to suicide attempts; there were unexplained numerical inconsistencies, which included a mathematical impossibility; there were unexplained exclusions of patients and data, and analyses were called intention-to-treat even though some patients with data were excluded; new outcomes appeared that were not prespecified in the trial protocol; rating scales and analyses were changed; the trial protocols were violated in other ways; and results that were inconsistent with the conclusion that fluoxetine is safe and effective were sidelined or explained away in a disturbing manner.
The efficacy outcomes were heavily biased in favour of fluoxetine by differential dropouts and missing data at trial endpoint. In trial X065, 6 patients had discontinued on fluoxetine and 12 on placebo after four weeks; in trial HCJE, none had dropped out on fluoxetine versus 10 on placebo after two weeks. Most analyses used the last observation carried forward (LOCF) method, but Lilly did not alert its readers to the large bias this caused: More patients on placebo than on fluoxetine had high depression scores carried forward.
FDA’s medical reviewer noted that considerably more patients on placebo than on fluoxetine dropped out and that the pattern was “rather unusual” in trial HCJE because there were more dropouts on placebo than on fluoxetine for adverse events (9 vs 5), patient decision (11 vs 3), and lost to follow up (7 vs 1).280 In contrast, trial X065 had no dropouts for adverse events on placebo versus 5 on fluoxetine, and there were no losses to follow-up.
Since no statistical adjustments can substitute missing data reliably, we focused on patients with minimal symptoms and those who had recovered after eight weeks in trial X065 according to Lilly’s own criteria. We did not find any differences to placebo.
In trial HCJE, Lilly’s analysis of the Clinical Global Impressions Efficacy Index, which compares the benefits and harms for each patient, was also misleading. The psychiatrists used an index with eight categories to “rate overall therapeutic effect in conjunction with side effects for each patient.” They assessed, for each patient, if the improvement in the depression outweighed any drug harms in terms of their interference with daily activities. This subjective process was not defined. Lilly claimed the results indicated that the therapeutic effects outweighed any harms because 58% vs 40% had a favourable score. However, when we combined the data from the eight categories by subtracting bad outcomes from good outcomes, which was more appropriate, we found that 59% versus 55% had a good outcome (P = 0.58).
Despite all the biases and manipulations we identified, the effects Lilly reported were not clinically relevant. The effect on the Children’s Depression Rating Scale—Revised, which is assessed by the psychiatrists or their research assistants, relative to the baseline values was only 4% in both trials for observed cases, and 16% and 9%, respectively, if LOCF is used. By comparison, the least recognisable effect on the equivalent adult scale, the Hamilton depression scale,267 corresponds to 28% of a median baseline of 25.4 in 35 placebo-controlled trials.269
It is more important what the patients think about the effect than what the psychiatrists and Eli Lilly think, and patient ratings did not find fluoxetine effective. Children’s Depression Inventory (CDI) was used for those below 13 years of age and Beck Depression Inventory (BDI) for those aged 13 and above. In trial X065, the data were combined in the publication and P = 0.58. In trial HCJE, the children even tended to prefer placebo: “Placebo-treated patients exhibited greater numerical reductions in the change from baseline for CDI and BDI total scores compared with fluoxetine-treated patients.”
Suicidal events were missing, both in the study reports and the publications. The published report for trial X065 did not mention that two patients on fluoxetine had attempted suicide, and the adverse events in four additional patients who discontinued fluoxetine were called “minimal,” even though three of them developed symptoms of mania and the fourth had a severe rash. Lilly’s internal report showed that 32 fluoxetine vs 18 placebo patients experienced at least one adverse event (P = 0.008), 19 vs 6 experienced restlessness (P = 0.005), 9 vs 1 had nightmares (P = 0.02), and 7 vs 4 felt tense inside. These are serious harms. Restlessness, including feeling tense inside, and nightmares increase the risk of suicide and violence.2,7
A subsequent publication by Lilly staff was also untrustworthy.318 It addressed safety in trial HCJE and had other numbers of suicidal events than those in the study report submitted to drug regulators.279,318
For trial HCJE, only the 9-week results were fully published, whereas the less positive 19-week results were not. Lilly falsely concluded that “fluoxetine 20 to 60 mg/day is safe” and also that “dose escalation may benefit some patients” even though they only reported on four outcomes for which there were no significant differences.
A 2007 Lilly meta-analysis of violent events, which included all placebo-controlled studies of fluoxetine undertaken in children and adolescents, was also untrustworthy.319 It is totally implausible that aggression or hostility-related events were experienced by fewer children and adolescents treated with fluoxetine, 2.1%, than by those treated with placebo, 3.1%.
Lilly’s results contradicted our findings and also FDA’s assessment of Lilly’s application. FDA created a table of discontinuations because of adverse events in X065, HCJE and HCJW (a trial of obsessive-compulsive disorder comparing fluoxetine 10-60 mg daily with placebo for 13 weeks in 71 vs 32 patients).280 There were 14 vs 3 discontinuations (P = 0.02, our calculation) among the 228 vs 190 patients for reasons related to suicide and violence (suicide attempt, euphoria, manic reaction, agitation, hyperkinesia, nervousness, personality disorder, hostility, and depression). In these trials, there were three suicide attempts on fluoxetine and one on placebo, and another fluoxetine patient was hospitalised because of suicidality. Six patients (2.6%) on fluoxetine developed mania or hypomania versus none on placebo (P = 0.03).280
In our restoration of trials X065 and HCJE, we found that precursors to suicidality or violence occurred more often on fluoxetine than on placebo. For trial HCJE, the number needed to harm was only 6 for nervous system events, 7 for moderate or severe harm, and 10 for severe harm. Fluoxetine reduced height and weight over 19 weeks by 1.0 cm and 1.1 kg, respectively, and prolonged the QT interval.
Lilly claimed in its study report for trial HCJE that “depression is an organic disease that readily responds to treatment” and that “Introduction of effective antidepressant treatments earlier in the progression of the disease state has the potential to effectively treat and control the disease as well as improve daily functioning and overall quality of life.”279 There is no evidence that either of these are true,7,8 and depression pills seems to worsen quality of life8 (see below).
If extrapolated from the trial data, the harm fluoxetine causes on growth in children corresponds to an annual loss in height of 2.7 cm and a loss in weight increase of 3.0 kg.279 The FDA requested that Lilly conduct a one-year study of the effect of fluoxetine on growth, which the company declined to do.280 We do not know if fluoxetine also has deleterious effects on the developing brain, but given what we know about psychoactive substances, including alcohol, this is likely.
Based on our reanalysis of the two pivotal trials, we concluded that fluoxetine is unsafe and ineffective.279 It is a horrible drug.
A textbook described the following treatment priorities for children with affective disorders: 1) psychoeducation and support; 2) cognitive behavioural therapy; 3) drugs.19:214 However, it also said that first-line therapy in severe depression is a combination of fluoxetine and cognitive behavioural therapy and that, for pronounced suicidal thoughts, hospital admission should be considered due to the risk of worsening suicidal thoughts and plans and a decrease in psycho-motor inhibition caused by increasing the drug dose.
It defies logic why an increase in dose is recommended in children who are suicidal when the authors acknowledge that this increases the risk of suicide and when we know that the pills do not even have a beneficial effect on the depression.
This is bad medicine but the lack of logic is ubiquitous. In New Zealand,8 the drug regulator did not approve the use of fluoxetine for people less than 18 years of age. However, this was no hindrance for the usage of depression pills, which increased by 78% between 2008 and 2016,320 and a UNICEF report from 2017 showed that New Zealand had the highest suicide rate in the world among teenagers between 15 and 19, twice higher than in Sweden and four times higher than in Denmark.321
I visited John Crawshaw, Director of Mental Health, Chief Psychiatrist and Chief Advisor to the Minister of Health in New Zealand, in February 2018, and I asked him to make it illegal to use these drugs in children to prevent some of the many suicides.8 He responded that some children were so severely depressed that depression pills should be tried. When I asked what the argument was for driving some of the most depressed children into suicide with pills that didn’t work for their depression, Crawshaw became uncomfortable, and the meeting ended soon after.
“Pushing children into suicide with happy pills” is the title for one of the chapters in my 2013 book about organised crime in the drug industry.6 Doctors cannot do worse than this: Telling children and their parents that happy pills are helpful when they don’t work and drive some children into suicide.
To see the list of all references cited, click here.
Mad in America hosts blogs by a diverse group of writers. These posts are designed to serve as a public forum for a discussion—broadly speaking—of psychiatry and its treatments. The opinions expressed are the writers’ own.
Mad in America has made some changes to the commenting process. You no longer need to login or create an account on our site to comment. The only information needed is your name, email and comment text. Comments made with an account prior to this change will remain visible on the site.