It is necessary to critically appraise research to ensure that only relevant information is discovered and considered evidence based. The following research article was selected to investigate the potential for serious harms associated with SSRI/SNRI antidepressants. It was a systematic review, meta-analysis, and quantitative designed study related to the following PICOT question: In adolescent and young adults ages 13-25 diagnosed with depression (P), does treatment with antidepressant medications (I) compared to an alternative therapy (C) result in an increase in violent and/or suicidal behaviors (O) over a period of twelve months (T)?
Why was the study done?
The purpose of this study was to evaluate the risk of suicide and aggression in children being treated with antidepressants. It was difficult for this research team to find published trial reports that included harms associated with antidepressants; therefore, they chose to analyze the clinical study reports conducted by drug companies for market approval. It was recommended, based on this study’s outcomes, that treatment with antidepressants in children and adolescents be avoided if possible because the risk of developing violent/aggressive behavior and/or suicide doubled (Sharma, Guski, Freund, & Gotzsche, 2016).
What is the sample size?
This study was a systematic meta-analysis review of approximately seventy double blind placebo-controlled trials. The researchers reviewed about 18,526 patient records that within the documentation reported harms experienced with taking SSRI or SNRI antidepressants. The harms included in the study had to be documented from patient narratives or provider listed harms per patient evaluation. The primary adverse effects under review were mortality, suicidality, violent behavior, and akathisia. Akathisia has been defined as a heightened state of restlessness and can increase the patient’s risk for suicide and violence (Sharma et al., 2016, Introduction). The cases examined were pulled from European and UK drug regulatory reports of five common SSRI antidepressants including: Cymbalta, Paxil, Zoloft, Prozac, and Effexor. In addition to this, some study reports were also obtained from Eli Lily’s website for Prozac and Cymbalta.
Are instruments of the variables in the study clearly defined and reliable?
This research study revealed 16 adult deaths and 155 suicidal events. Placebo results showed mortality in adults with an odds ratio 1.28, 95% CI 0.4-4.06, suicidality in adults 1.21, CI 0.84-1.74, and akathisia in adults 2.04, CI 0.93-4.48. Comparatively, the results of adults taking antidepressants were significant in that they showed more aggression at 1.93, CI 1.26-2.95, suicidality 0.81, CI 0.51-1.28, akathisia 2.00, CI 0.79-5.04. The results in children/adolescents were of highest statistical significance with a 2.39, CI 1.31-4.33 for increased aggression, 2.79, CI 1.62-4.81 for suicidality, and 2.15, CI 0.48-9.65 for akathisia. In the study reports collected from Eli Lilly, all mortality was accurately reported, all suicidality related events were omitted, and all the remaining harms related outcomes were incomplete and therefore unreliable. The study results did indicate to the research team that there was no significant rise in adult aggression and suicidal outcomes but the risk of these harmful events in the youth age population, doubled. The risk for akathisia in both adults and youth age participants doubled (Sharma et al., 2016, Results). It was noted that because of the shortcomings identified and having only partial access to case study report forms, the harms could not be estimated as accurately. To increase the reliability of investigated harms in this study, access to anonymized individual patient data was needed.
How was the data analyzed?
Two researchers extracted data independently; the outcomes were meta-analyzed by Peto’s exact method, also referred to as Peto’s odds ratio or fixed effect model.
Were there any unusual events during the study?
Sharma et al. (2016) discovered during their evaluation of these trials, that harmful events were not accurately reported. The researchers discovered that many trials resulting in aggressive behavior outcomes did include this data in the study reports, unlike that of suicidality, but the ball had been dropped on reporting safety concerns as they related to violent crime. The investigative team found this concerning since many of the killers involved in school shootings and mass homicidal events had been taking psychotropic medications and were ruled not guilty due to drug induced insanity (Gotsche, 2015, Chapter 14). In their results, they also disclosed that four deaths were negligently reported by the drug company and documentation altered favorably for the antidepressant. Documentation language was also worded to label harmful symptoms as the result of worsening mental health status and not a potential adverse effect of the anti-depressant drug. In addition, personal patient accountings were found to be withheld completely, raising another red flag that not all adverse or harmful events were being fully disclosed.
How do the results fit in with previous research in this area?
Results, that suicide and aggression as they are impacted by antidepressant therapy, were found to be inconclusive upon completion of this thorough review. This was due to low incidence of adverse events, poor study design, and inaccurate and biased trial reporting. The researchers concluded that there is a significant call for improved clinical reporting measures that are scientifically and ethically sound. This is a common theme found in a lot of the previous research surrounding this topic. The results were inconclusive, bias in reporting, inaccurate reporting, and more reliable research data is needed. For the benefits and harms of drug treatment therapies to most accurately be investigated to determine safest and best outcomes for science and humanity, these unreliable reporting practices need to change. If the primary goal of scientific discovery is the acquisition of truth, then these biased research methods are an insult to science (Sharma et al., 2016).
What are the implications of the research for clinical practice?
Based on the outcomes generated from this research, it is suggested that providers should limit use of antidepressants in clinical practice to treat depression in children, adolescents, and young adults as serious harms appear to be higher within this population group. Alternative anti-depressant therapies such as psychotherapy, exercise, and dietary/supplemental changes, may have some potential benefits in youth age patients and should be considered. The FDA has advised that antidepressants may increase suicidality in young adults ages 18-24 and recommend that patients of all ages treated with antidepressants be monitored for clinical worsening, suicidality, and unusual or harmful changes in behavior (Ho, 2012).