Study's conclusion is unwarrantedBMJ 1995; 311 doi: https://doi.org/10.1136/bmj.311.6996.56 (Published 01 July 1995) Cite this as: BMJ 1995;311:56
- Morris Bernadt,
- Rachel Hammill
EDITOR,--We believe that one of the key messages in the paper by John A Henry and colleagues is misleading: it states that “deaths from overdose can be prevented by switching prescribing from older, tricyclic drugs; such a move may contribute to government targets for reducing suicide.”1 The same point is made in the last paragraph of the paper. The flaw is to equate the prevention of fatal poisoning with that of suicide in general. In the same issue Susan S Jick and colleagues found that of 143 suicides in Britain, only eight (6%) were due to antidepressant overdose.2 Similarly, in Sweden 190 (6%) of 3400 people who committed suicide had toxic blood antidepressant concentrations.3 The figure for England and Wales in 1990 was 7%. Thus the overwhelming majority of suicidal acts involve means other than an antidepressant overdose. Other things being equal, the strategy of prescribing newer, non-toxic antidepressants would have at best a marginal effect on the suicide rate.
Newer antidepressants may not, however, be as efficacious in treating depression as older ones.3 In a meta-analysis selective serotonin reuptake inhibitors were significantly less efficacious than tricyclic antidepressants on the 17 item Hamilton depression rating scale, but the difference was clinically small.4 Of 16 trials used for pooled analysis, two showed an advantage for tricyclic antidepressants, 14 showed no difference between selective serotonin reuptake inhibitors and tricyclic antidepressants, and none showed an advantage for selective serotonin reuptake inhibitors. Since then nortriptyline has been shown to be much more efficacious than fluoxetine in elderly depressed patients.5
When all causes of suicide are considered some of the newer antidepressants are associated with a higher overall suicide rate--for example, the adjusted relative risk for fluoxetine was 2.1 times that for dothiepin2 and the standardised mortality ratio for suicide associated with mianserin was twice that associated with amitriptyline.3 Authors of both studies addressed the issue of biases that might have arisen because of non-random allocation of patients to antidepressant type. We think, however, that the conclusion that one might reduce the suicide rate by preferentially prescribing the newer antidepressant drugs is unwarranted.