In February of 2018 the international debate on antidepressants was renewed when James Davies, a co-founder of the Council for Evidence-Based Psychiatry (CEP), and his coauthors published a letter in the Times on the benefits and harms of antidepressants. This was in response to a study done by Cipriani et al that found all the 21 antidepressants reviewed to be more effective than placebo. Carmine Pariante of the Royal College of Psychiatrists said: “This meta-analysis finally puts to bed the controversy on anti-depressants, clearly showing that these drugs do work in lifting mood and helping most people with depression.” In response, the Council for Evidence-Based Psychiatry said that statement was “irresponsible and unsubstantiated, as the study actually supports what has been known for a long time,” namely that the differences between placebo and antidepressant are so minor that they are clinically insignificant. It created a media and professional firestorm that has yet to burn out, and even led to some strategic retreats by organizations like the RCP that originally hailed the results.
CEP noted how the individuals in the referenced studies were not in truly blinded clinical trials. “Most people on antidepressants experience some noticeable physical or mental alterations, and as a consequence realise they are on the active drug.” This then boosts the placebo effect, adding further questions about the so-called effectiveness of antidepressants. Irving Kirsch has published several studies demonstrating the significance of the placebo effect with antidepressants. For more on the Cipraini et al study, see “The Lancet Story on Antidepressants,” Part 1 and Part 2. For more on Irving Kirsch and the placebo effect, see “Dirty Little Secret.”
Additionally, the trials only addressed short-term use of antidepressants (8 weeks), not the long-term use which is more typical. “Around 50% of patients have been taking antidepressants for more than two years, and the study tells us nothing about their effects over the long term. In fact, there is no evidence that long-term use has any benefits, and in real-world trials (STAR-D study) outcomes are very poor.” STAR*D was the largest, longest and most expensive study of antidepressants ever conducted.
James Davies and John Read (also a member of CEP) published a systematic review in the journal Addictive Behaviors that showed antidepressant withdrawal was “more widespread, severe and long-lasting than indicated by current guidelines.” The review indicated that an average of 56% of patients who stop or reduce their antidepressants experience withdrawal symptoms, a significant proportion of whom experienced them for more than two weeks. “It is not uncommon for patients to experience symptoms for several weeks, months, or longer.” One study said 40% of patients experience symptoms for at least six weeks; another indicated that 25% experience symptoms for at least 3 months. Davies said the new review indicated what patients have known for years, “That withdrawal from antidepressants often causes severe, debilitating symptoms which can last for weeks, months or longer.”
Davies and Read noted in their paper that an implication of the higher incidence of antidepressant withdrawal and longer duration added credence to concerns that doctors were misdiagnosing antidepressant withdrawal as treatment failure. “Re-emergent symptoms of depression and anxiety are a regular feature of antidepressant withdrawal itself.” They pointed out where the RCP’s own survey, “Coming Off Antidepressants” found that the withdrawal reaction was rated severe by most people, and approximately 25% of users reported experiencing anxiety for at least 3 months after stopping their antidepressant.
The President of the Royal College of Psychiatrists, Wendy Burn, published a letter in the Times that said “We know that in the vast majority of patients, any unpleasant symptoms experienced on discontinuing antidepressants have resolved within two weeks of stopping treatment.” CEP challenged the Royal College of Psychiatrists and its president, stating they believed the statement was not evidence-based; that it misled the public. Further, they pointed out how within 48 hours of the misleading statement in the Times, the RCP removed “Coming Off Antidepressants” from its website. They suggested one interpretation of that action was the RCP was attempting keep the public from seeing evidence that contradicted what the RCP president claimed in the Times.
This was not just a dispute between CEP and the RCP over interpreting Cipriani et al. August of 2018 contained a one-two punch that broadened the debate over antidepressant ineffectiveness. The British Journal of Psychiatry published an editorial written by Gordon Parker, the founder of The Black Dog Institute, “The benefits of antidepressants: news or fake news?” that said antidepressant trials were disconnected from the real world of clinical practice. Psychological Medicine published a study by de Vries et al that analyzed the cumulative effect of publication biases on the apparent efficacy of antidepressants for the treatment of depression.
Asking if antidepressants are effective treatment for major depression is asking the wrong question. The problem, according to Gordon Parker, is that ‘major depression’ is a “domain diagnosis” for a variety of depressive illnesses. “Basically, the target diagnosis of major depression captures multiple types of depressions—some biological, some psychological, some social—and not all would be expected to respond to medication.” In other words, you lose the evidence for their effectiveness with biological causes by combining them with social and psychological ones. “For patients with depression, if you narrow down to those who have a biologically-based depressive sub-type, the antidepressants are distinctly effective.”
De Vries et al looked at the cumulative impact of biases upon on two effective treatments for depression: antidepressants and psychotherapy. They identified four major biases: study publication bias, outcome reporting bias, spin, and citation bias. Study publication bias involves not publishing an entire study. Outcome reporting bias refers to not publishing negative outcomes or switching the status of primary and secondary outcomes. “Both biases pose an important threat to the validity of meta-analyses.”
Spin uses reporting strategies that distort the interpretation of results and mislead readers. Authors conclude the treatment is effective despite non-significant results on the primary outcome. For example, by focusing on statistical significance instead of clinical significance, researchers have confirmed the efficacy of several SSRIs. Another spin technique is instead of concluding a treatment was no more effective than placebo, researchers point out how a treatment was well tolerated and effective in a sub population of the original study, say patients who had not received prior therapy. Finally, with citation bias, studies with positive results receive more citations than negative studies. This leads to greater visibility of positive results and creates an obstacle to ensuring that negative findings can be discovered. De Vries et al concluded:
The problem of study publication bias is well-known. Our examination of antidepressant trials, however, shows the pernicious cumulative effect of additional reporting and citation biases, which together eliminated most negative results from the anti-depressant literature and left the few published negative results difficult to discover. These biases are unlikely to be unique to anti-depressant trials. We have already shown that similar processes, though more difficult to assess, occur within the psychotherapy literature, and it seems likely that the effect of these biases accumulates whenever they are present. Consequently, researchers and clinicians across medical fields must be aware of the potential for bias to distort apparent treatment efficacy, which poses a threat to the practice of evidence-based medicine.
In October of 2018 a reanalysis of the STAR*D study, supported the claim of antidepressant ineffectiveness. The STAR*D study, published in 2004, attempted to mimic real world patients, recruiting from routine outpatient treatment centers. Additionally, they did not exclude patients with comorbid diagnoses, as is typically cone in clinical trials. STAR*D was funded by the NIMH at a cost of $35 million dollars and took six years to complete. The reanalysis was done by Irving Kirsch and others. The improvement found in the reanalysis was roughly half of that seen in the standard comparative drug trials. In her review of the Kirsch-led reanalysis for Mad in America, Joanna Moncrieff said STAR*D suggested that “in real life situations (which the STAR-D mimicked better than other trials) people taking antidepressants do not do very well.”
For the vast majority of people, depression naturally remits. “It is difficult to believe that people treated with antidepressants do any better than people who are offered no treatment at all.” Moncrieff speculated this may be the reason why the results of the main outcome of the STAR*D study took so long to be published. For more on the STAR*D study, see “Antidepressant Fall from Grace, Part 2.”
Then in May of 2019, the Royal College of Psychiatrists changed its position on antidepressant withdrawal. It issued a revised policy statement updating its guidance to doctors. James Davies of CEP said the changes were welcome; and if acted upon, “will help reduce the harm that is being caused to huge numbers of patients through overprescribing, inadequate doctor training and often disastrous withdrawal management.” The College called for the following changes:
- There should be greater recognition of the potential for severe and long-lasting withdrawal symptoms on and after stopping antidepressants in NICE guidelines and patient information
- NICE should develop clear evidence-based and pharmacologically-informed recommendations to help guide gradual withdrawal from antidepressant use
- The use of antidepressants should always be underpinned by a discussion with the patient about the potential level of benefits and harms, including withdrawal
- Discontinuation of antidepressants should involve the dosage being tapered, which may occur over several months, and at a reduction rate that is tolerable for the patient
- Monitoring is needed to distinguish the features of antidepressant withdrawal from emerging symptoms
- Adequate support services should be commissioned for people affected by severe and prolonged antidepressant withdrawal, modelled on existing best practice
- There should be routine monitoring on when and why patients are prescribed antidepressants
- Training for doctors should be provided on appropriate withdrawal management
- Research is needed into the benefits and harms of long-term antidepressant use
These changes by the RCP with regard to antidepressants are needed in the US as well. Antidepressant withdrawal is a real concern for some individuals. Routine monitoring of when and why patients are prescribed antidepressants is needed. Support services are needed for individuals who experience severe and prolonged withdrawal. There is a need to inform patients when prescribing antidepressants of the potential benefits as well as the potential harms—including withdrawal.
Research into the potential benefits and harms of long-term antidepressant use is needed. Discontinuation of antidepressants should be done slowly, taking its cue from how well the patient is tolerating the taper. Both the patient and doctor should carefully monitor the tapering process and strive to distinguish between symptoms of antidepressant withdrawal and emerging symptoms of the underlying depressive disorder. Doctors need to be trained in appropriate tapering and withdrawal management of antidepressants.
Drawing on the above discussion, we can add the need for greater awareness of the multiple types of depressions—some biological, some psychological, some social—and the need to freely acknowledge that antidepressants won’t work for everyone. Edward Shorter makes a compelling case for distinguishing between depression and melancholia in How Everyone Became Depressed. In the pursuit of developing the evidence base for the use of antidepressants and best practice guidelines, we need to systematically eliminate the impact of bias on the publication of research results with antidepressants. Admittedly this is a problem that extends beyond just antidepressant research, see “Clinical Trial Sleight-of-Hand,” “The Reproducibility Problem” and “Reproducibility in Science” for more information.
British psychiatrists have taken the first step towards correcting errors in how they use antidepressants. Hopefully they will persist in seeing that the recommended changes are implemented. American psychiatrists and physicians need to do the same. They need to follow the lead of the RCP.