Key systematic reviews of the evidence
This review examined 97 studies over the past 50 years on the imitative effect of presentations of suicide in newspapers, television, books and the internet. It specifically considered methodological issues associated with each body of studies, and critically examined whether those issues limited the inferences that could be made about the interpretation of findings.
The review examined the strength of the association between the portrayal of suicide in the media and actual suicidal behaviour, and the extent to which that association could be considered causal. In determining causality, it examined five features of the findings:
- Consistency: whether the association was consistently observed, regardless of study design and population sampled.
- Strength: whether the association was statistically significant, and particularly whether there was evidence that the greater the exposure to the media coverage of suicide, the greater the increase in suicide rates.
- Temporality: whether exposure to the media coverage of suicide occurred before actual suicides.
- Specificity: whether the people exhibiting the suicidal behaviour were actually exposed to media coverage of suicide.
- Coherence: whether the association was in line with known facts about the influence of media and suicide.
The review examined 41 studies of differing types. The vast majority suggested an association between newspaper reports of suicide and actual suicidal behaviour. From that body of evidence the review found that it was reasonable to regard the association was causal, because:
- the association was reliably observed under almost all study methodologies;
- a dose‐response effect was evident such that the greater the newspaper coverage of a particular suicide, the more substantial the increase in subsequent suicides; and
- the findings were coherent in that they made sense in light of what is already known about the influence of the media and suicide.
Despite the finding of causality, the review noted that only a limited number of studies permitted a determination of whether the media stimuli preceded an increase in suicide rates, and only a few studies were able to demonstrate that a reasonable proportion of those who subsequently died by suicide were exposed to the media stimulus.
The review examined 11 studies about television (all of a similar design), 20 studies of the internet (mostly descriptive studies), and 20 studies of mixed media (of differing designs). The review found that the studies provided cautious support for a causal association:
- The associations were generally consistent. With only a few exceptions, the studies all found a similar association.
- In each case, the findings were coherent.
- The internet and mixed media studies were able to show the temporality and specificity of the association, with strong evidence that individuals were exposed to the suicide material before their suicidal behaviour. The television studies based their data on monthly rather than daily figures so were not able to demonstrate that the media reports actually occurred before the suicidal behaviour, nor that the person was exposed to the report.
- The television studies demonstrated the strength of the association in that it was most evident immediately after the media stimulus and then dissipated. The internet and mixed media studies were not designed to be able to demonstrate the strength of the association.
The review also examined five studies about books, but they were all about the book Final Exit, which advocates suicide for the terminally ill. It found that the association between that book and suicidal behaviour was causal.
The purpose of this review of 56 studies was to give an overview of the research and to find out the effects of media reporting of suicide on actual suicidality. The review pulled 1180 studies from electronic databases and whittled them down to 56 eligible analytical research reports on non-fictional portrayals of suicide behaviour. The review found that the vast majority of studies support the idea that media coverage of suicidal behaviours and actual suicidality are associated. Only four studies found no significant associations (all before 1990) and a further five expressed hesitations about clear associations or reported incoherent results. The review warned, however, of a risk of reporting bias in that “researchers are eager to report meaningful positive results, but could keep silent if the results are not beneficial”.
This study was a meta-analysis of 10 previous ecological studies covering 98 celebrity suicides. This is a particularly strong study design as it systematically re-analysed the published aggregate data from the previous studies and provided cumulative insights into the association between media reports of celebrity suicides and subsequent suicides.
This study found that:
- The average increase in the suicide rate across the pooled results was 0.26 in the month after a report on a celebrity suicide (with 95 per cent of the results lying between 0.09 and 0.43).
- The imitative effect was most pronounced after reports on suicides by national top entertainers.
- There was a substantial diversity of results across the studies, which could be explained by regional differences in reporting practices for suicides.
This review suggested that the variation in findings across the studies examining the impacts of media coverage of suicide may be explained by their different methodologies. The review examined 419 findings from 55 different studies. It found:
- Substantial variations in the effects found by different studies.
- Studies that examined reports of a suicide by an entertainment or political celebrity were 5.27 times more likely than other studies to find a copycat effect.
- Studies based on stories with powerful negative definitions of suicide were 99 per cent less likely than other research to report a copycat effect. Examples of suicide stories that stressed negative definitions were Kurt Cobain’s death (which included condemnation of the suicide by his spouse) and the 1978 mass suicide at Jonestown (where the coverage included footage of rotting bodies in piles and quickly labelled the suicides as cult-inspired).
- Studies based on television coverage were 79 per cent less likely to report an imitative effect than studies based on newspaper stories.
- Studies that examined female suicide rates were 4.89 times more likely to find copycat effects than those based on male rates or total suicide rates.
- Studies examining the suicidal behaviour of young people were 54 per cent less likely to find an imitative effect than studies examining the total population.
- Of the 419 findings reported in the 55 studies, 35.8 per cent found an imitative effect.
This narrative review reported on the range of studies examining the association between suicide reporting and subsequent suicidal behaviour that had been conducted to date. It placed particular emphasis on identifying the different characteristics of non-fictional reporting or of the vulnerable person that influence the size of the contagion effect. It found:
- The studies clearly demonstrate that extensive newspaper coverage of suicide is associated with a significant increase in the rate of suicide, and that the magnitude of the increase is proportional to the amount, duration and prominence of media coverage.
- There is overwhelming evidence that the magnitude of the increase in suicidal behaviour after newspaper coverage of a suicide is related to the amount of publicity given to the story and the prominence of the story in the newspaper.
- Studies found that celebrity status of the person reported to have died by suicide had a greater impact than non-celebrity stories.
- Studies found that the imitative effect is greatest for teenagers.
- Interestingly, the imitative effect does not appear to be stronger for demographic groups already strongly predisposed to suicide (“males, whites, unmarried people, or retired persons”).
- There is increasing research evidence to support the suggestion that the imitative effect is greater amongst individuals who share similar life and demographic characteristics to the person reported to have died by suicide.