Venlafaxine for major depressionBMJ 2007; 334 doi: http://dx.doi.org/10.1136/bmj.39098.457720.BE (Published 01 February 2007) Cite this as: BMJ 2007;334:215
- Andrea Cipriani, lecturer in psychiatry ()1,
- John R Geddes, professor of epidemiological psychiatry2,
- Corrado Barbui, lecturer in psychiatry3
- 1Department of Medicine and Public Health, Section of Psychiatry and Clinical Psychology, University of Verona, 37134 Verona, Italy
- 2Department of Psychiatry, University of Oxford, Oxford OX3 7JX
- 3Department of Medicine and Public Health, Section of Psychiatry and Clinical Psychology, University of Verona
In this week's BMJ Rubino and colleagues provide new evidence that informs the debate about whether antidepressants increase the risk of suicide.1 The study, a retrospective observational analysis of the General Practice Research Database, found that patients prescribed venlafaxine were more likely to attempt or complete suicide than patients prescribed citalopram, fluoxetine, or dothiepin. Adjustment for possible confounders, however, greatly reduced the excess risk.
Venlafaxine is a serotoninergic and noradrenergic reuptake inhibitor, and it may be more effective than selective serotonin inhibitors for major depression.2 3 However, patients often discontinue treatment because of side effects.4
The database analysed is the world's largest computerised database of anonymised longitudinal medical records from primary care (more than 3.4 million active patients, about 13 million in total since 1987, from around 450 primary care practices throughout the United Kingdom; www.gprd.com). The database has been used extensively for pharmacoepidemiological research, including previous studies examining the possible association between death from suicide and the use of antidepressant drugs.5 6 7
The non-experimental nature of the database creates methodological problems that can make results difficult to interpret, as the study by Rubino and colleagues shows. Firstly, the patients treated with venlafaxine were probably selected clinically, and differed from those treated with other agents in several variables related to the risk for suicide. Although sophisticated statistical analyses were used to control for potential confounding, some variables—such as diagnosis, comorbidities, and pre-existing depression or suicidal ideation—may not have been effectively accounted for. Moreover, there may have been residual confounding by uncontrolled variables, such as treatment dosages and adherence to treatment.8
Secondly, the choice of the primary outcome is important. Although some epidemiological studies have used the outcome of all deliberate self harm, Rubino and colleagues restricted the primary outcome to acts with a deliberate suicidal intent. Deliberate self harm, particularly suicide, is often thought to be a relatively “hard” outcome in studies of antidepressants, but enormous scope exists for ascertainment bias. For this reason, in a meta-analysis of randomised clinical trials of long term lithium therapy, we used all-cause mortality as the primary outcome, and suicide and deliberate self harm as secondary outcomes, to limit ascertainment bias and make the findings more robust.9
Thirdly, differences in the drug being compared and sample populations may explain heterogeneity between the results of different observational analyses.6 10 For example, Rubino and colleagues selected only specific antidepressants as comparators (citalopram, fluoxetine, and dothiepin), excluded other agents, and did not include a reference group of patients not taking antidepressants.
Observational evidence offers insights into long term and real world outcomes for large groups of people, but it can rarely show a convincing causal relation between two events. It can be hypothesised that the drug itself can precipitate suicide, because of its potential mechanism of action. Randomised controlled trials are better able to establish causal relations, but they usually follow highly selected samples of patients for short periods.11 Systematic reviews of randomised controlled trials may increase statistical power, but absolute numbers of patients having rare adverse events such as completed or attempted suicide are low. Thus, reporting or not reporting a few cases can completely change the overall outcome.12 Even with these limitations, systematic reviews have consistently reported an excess risk of suicide in children and adolescents with major depression taking antidepressants, but not in adult patients.13 14 15 16
Despite these uncertainties, clinically useful conclusions for everyday clinical practice need to be formulated. Currently, UK guidelines recommend that treatment with venlafaxine should be started or managed only under the supervision of specialist mental health medical practitioners.17 The Medicines and Healthcare Products Regulatory Agency has recently changed this guidance, however, to apply only to severely depressed patients or those in hospital who need doses of 300 mg daily.18
Recent observational evidence indicates that, in suicidal patients who have ever used antidepressants, current use of any antidepressant is associated with an increased risk of attempted suicide and with a decreased risk of completed suicide and death.10 In this analysis, venlafaxine was associated with the highest risk of suicide.10 A similar risk profile for venlafaxine was highlighted by reanalyses of data from clinical trials conducted in children.14 Finally, the Food and Drug Administration in the United States, the Medicines and Healthcare Products Regulatory Agency, and the manufacturer of venlafaxine have issued a warning about the risk of cardiotoxicity and toxicity associated with overdose of venlafaxine.19 20
Venlafaxine therefore has a consistent but unexplained risk of increased suicide and toxicity. Despite evidence of its marginally greater efficacy compared with other antidepressants,2 3 current evidence suggests that venlafaxine should not be first line treatment for people with major depression.
Competing interests: JRG has received research funding and support from GlaxoSmithKline, Sanofi-Aventis, UK Government Department of Health, UK Medical Research Council, and the Stanley Medical Research Institute.