New evidence from clinical study reports reveals misclassification, misrepresentation, and under-reporting of serious harm
People in the United Kingdom are consuming more than four times as many antidepressants as they did two decades ago.1 Despite this, we still do not fully understand the effects of these drugs, nor can we be confident about their risk-benefit ratio.
Unresolved problems include the nature and frequency of potential serious adverse effects such as self harming or suicidal actions, suicidal ideation, and aggression. In the linked paper, Sharma and colleagues use clinical study reports to explore such adverse effects.2 Clinical study reports are prepared by pharmaceutical companies for obtaining marketing authorisation. Guidelines on the contents and presentation of clinical study reports exist,3 but requirements are not mandatory. Consequently, although clinical study reports usually contain more data than published articles,4 the level of detail varies. Moreover, what is reported in the body of the clinical study reports may differ from data in appendices, such as individual patient listings of adverse events or narratives of serious adverse events.
The analysis of clinical study reports by Sharma and colleagues reveals misrepresentation of adverse events. Comparing the “results” reported in the reports with data from individual patient listings or patient narratives revealed misclassification of deaths in people taking antidepressants and misrepresentation of suicidal events. More than half of the suicide attempts and instances of suicidal ideation were coded as “emotional lability” or “worsening of depression,” for example. Summary reports published on Eli Lilly’s website were even more incomplete, listing only 10% of the suicide attempts revealed in the corresponding clinical study reports, and no instances of suicidal ideation.
Over half of the clinical study reports selected by Sharma and colleagues had no individual patient listings, these data having been withheld. This begs the question of how many more adverse events would have been revealed if individual patient listings were available for all trials, and raises concerns about why this information is allowed to be withheld.
The current review is consistent with other evidence pointing to an increase in suicidal behaviour and ideation among children and adolescents taking selective serotonin reuptake inhibitors (SSRIs).5 The paper by Sharma and colleagues is the first large scale, quantitative analysis to show an increase in aggressive behaviour in this age group. The analysis did not detect an increase in suicide, suicide attempts, suicidal ideation, or aggression in adults, and meta-analyses of data on adults remain conflicting.67 Early case reports, published before the issue became controversial, document occurrence of these effects in adults, however, and provide evidence of the mechanism of these effects. These reports describe a state of agitation, somewhat similar to the akathisia produced by antipsychotics, associated with intense and violent suicidal preoccupation.8 However, as Sharma and colleagues show, recording of “akathisia” in antidepressant trials remains inconsistent and unreliable, hence we have little idea of the frequency of this effect.
Although “activation” effects (or akathisia) are recognised, there is little acknowledgment that antidepressants have mind-altering properties that are independent of their supposed effects on underlying mental disorders. One study revealed various changes to emotions and behaviour in patients taking antidepressants that were associated with suicidal ideation.9 Evidence suggests that withdrawal syndromes can be severe and sometimes prolonged.10 With some exceptions,11 however, we lack the detailed exploration through animal behaviour and healthy volunteer studies that might clarify the range, nature, and duration of the alterations that antidepressants induce.
Despite their widespread use, antidepressants are only modestly more effective than placebo in trials of depression. Measures of global clinical improvement suggest the difference is not clinically relevant or even detectable.12 Moreover, placebo controlled studies do not distinguish whether the effects of a drug are attributable to the targeting of putative underlying mechanisms or a consequence of the drug’s mind-altering effects. The blunting of emotions produced by SSRIs may directly affect depression rating scale scores, for example, and their psychoactive and physical effects may influence patient expectations, promoting an amplified placebo effect.13
With doubts about the efficacy and effectiveness of antidepressants and evidence that they can produce such serious adverse reactions as suicidal and aggressive tendencies, regulators and the public need access to more comprehensive and reliable data. The results reported in clinical study reports, on which decisions about market authorisation are based, are likely to underestimate the extent of drug related harms. We need access to original data from trials, but we also need more research to describe the whole range of antidepressant induced behavioural, emotional, and physical alterations with acute treatment, long term use, and withdrawal.
Competing interests: I have read and understood the BMJ policy on declaration of interests and declare the following: I have no competing financial interests. I am co-chair of the Critical Psychiatry Network.
Provenance and peer review: Commissioned; not externally peer reviewed.
Sharma T, Guski LS, Freund N, Gøtzsche PC. Suicidality and aggression during antidepressant treatment: systematic review and meta-analyses based on clinical study reports. BMJ 2016;352:i65.
International Conference on Harmonisation. Structure and content of clinical study reports, E3. Report No. E3. ICH, 1996.
Wieseler B, Wolfram N, McGauran N, et al. Completeness of reporting of patient-relevant clinical trial outcomes: comparison of unpublished clinical study reports with publicly available data. PLoS Med 2013;10:e1001526.
Whittington CJ, Kendall T, Fonagy P, Cottrell D, Cotgrove A, Boddington E. Selective serotonin reuptake inhibitors in childhood depression: systematic review of published versus unpublished data. Lancet 2004;363:1341-5.