Reporting suicide The effect of media coverage on patterns of self harmBMJ 1994; 308 doi: https://doi.org/10.1136/bmj.308.6941.1446 (Published 28 May 1994) Cite this as: BMJ 1994;308:1446
- D Gunnell
Suicide accounts for only 1% of all deaths and yet when these occur they frequently attract disproportionate media interest. Headlines such as “Coroner told of Human Torch Horror” and “ Tragic Countess of Caithness died from a single shot” are commonplace following the suicides of individuals whose death from other causes would probably been been left unreported.
William Farr wrote in 1841 that “no fact is better established in science than that suicide (and murder may perhaps be added) is often committed from imitation. . . . Do the advantages of publicity counterbalance the evils attendant on one such death? Why should cases of suicide be recorded at length in the papers any more than cases of fever?” Farr called for a cessation of the reporting of “dramatic tales of suicide, murder and bloodshed.”
There is clear evidence that the media may affect method specific suicide rates. In Britain an excess of about 60 suicides by burning occurred in the 12 months after the widely publicised political suicide by burning of a woman in Geneva.1 The evidence concerning the media's influence on overall suicide rates is less clear. Increase in suicide rates, following the reporting of real life suicide, have been described both in Britain and the United States. 2,3 The methodologies for these and other studies are questionable, with the examination of changes in rates over apparently arbitrarily selected periods of time and a failure to show overall increases over linger tome periods. Schmidtke and Hafner have, however, produced more robust evidence by examining suicide rates after two separate broadcasts of the fictional portrayal of a young man's suicide on a railway line. An imitation effect leading to methods specific and absolute increases in the number of suicides was seen. The imitation effects were greatest in those of the same age and sex as the fictional character, and the numbers of suicides closely corresponded with the audience figures for the two broadcasts. Effects were observed for up to 70 days after the broadcast; an estimated overall excess of 60 suicides occurred.4 The effect on suicide rates of a television series dramatising the work of the Samaritans has also been studied. Although the series led to a rise in new client referrals no effect was seen on the number of suicides.5
It is argued that suicides occur only among those who are already suicidal and it only the choice of method that is influenced by publicity. A century ago Durkheim argued that although media attention may precipitate clusters of suicide, these occur only among those who would commit suicide sooner or later any way, the publicity merely acting as a precipitant to an inevitable event. Schmidtke and Hafner observed greatest increases among those most similar to the “model” portrayed, but Ashton and Donnan did not. Disentangling these conflicting hypotheses is difficult as suicide is a rare event and the particular methods examined constitute only a small fraction of all suicide. In any event the effect is likely to be small.
The media and parasuicide
Parasuicide occurs more frequently, and the dramatisation of this too has been closely studied. The BBC series Eastenders and Casualty have both portrayed overdoses. The Eastenders broadcast in February 1986 was watched by over 14 million viewers and has been the subject of much debate. The most detailed study looked at admission for overdose in 63 hospitals in the days after the broadcast. Although rises in overdoses in women were seen, these did not occur in the expected age group and the author concluded that association remained unproved.6 Several individual accident and emergency departments reported increases in rates following the broadcast. Further analysis of one of these hospitals' finding showed that overdose attendences had been increasing for two months before the broadcast, and although attendances were higher than in previous years, the episodes could not have been responsible for this rise.7 When the characteristics of 34 of those seen after the episode were compared with a similar number seen over a month later, those seen following the episode were more likely not to have overdoses in the past and more often took analgesics, as did the fictional character.
Showing parasuicide on television might be beneficial. As a result of knowledge gained from an episode of Casualty highlighting the dangers of paracetamol overdose, a young girl was convinced by a friend of the need for (possibly lifesaving) hospital treatment.8
Guidelines on suicide, such as those produced by the BBC, advise against the reporting of “graphic details of method,” particularly when the method used is unusual. Care is also advised in the dramatised glorification of suicidal behaviour and actions. Only one study, however, has analysed the effects of introducing reporting guidelines on suicide. Etzerdofer et al reported an increasing trend in suicide and parasuicide on the Viennese underground system: 13 suicides occurred on the system in 1986 and nine in the first six months of 1987, there having been only nine suicides in 1980-4. Dramatic reporting of these suicides was thought to be partially responsible for this trend. Introducing reported guidelines improved the quality of reporting and sharply reduced the number of suicides on the underground (six in 1985, three in 1989, and four in 1990).9 Again, it not possible to determine this from study whether those who would have committed suicide on the underground chose an alternative method, as underground suicides contributed only a small proportion to overall suicide numbers in Vienna. However, if reducing access to a particular method effects both method specific and overall suicide rates controls on media reports may influence suicide rates in the same way.
In summary, it seems likely that media portrayal of both fictional and real life suicides may in some circumstances leads to small increases in overall suicide rates and larger increases in method specific suicide rates. This has not been in all studies and this effect may differ depending on the setting of the portrayal of the act, the method used, and other factors. A greater understanding of the features of the publicity liable to result in suicidal behaviour is needed. Guidance on the reporting of suicide by the media may decrease imitative suicides, which mat account for 1% of suicides annually. Etzerdorfer et al have shown that media guidelines do have an effect, and their findings deserve to be taken seriously.
The fact that patterns of suicide may be affected by portrayal in the media of these events suggests that other behaviours may be similarly influenced. This would support the views of those who call for greater restrictions and controls on the depiction of violence on our television screens. There is generally, however, a responsible attitude towards the reporting of suicide by most of the media, and legislative control in Britain is stricter than in many parts of the world. The potential benefits of promoting mental health by the media must not be forgotten as here too the media provide a powerful tool for the public health