Can news reporting of suicide raise or lower suicide deaths?

What happens after news of a celebrity's suicide?

A 2020 meta-analysis in the BMJ (31 studies, 20 in the main analysis; search up to September 2019) found that suicide risk rose by 13% after media reports of a celebrity's suicide (relative risk, RR — how many times higher the risk was: 1.13, 95% CI 1.08–1.18; 14 studies). The median follow-up was 28 days, ranging from 7 to 60. All the studies were observational — interrupted time series and before-and-after comparisons — so this is an association, not proof of cause.

When the method by which the celebrity died was reported, deaths by the same method rose by 30% (RR 1.30, 1.18–1.44; 11 studies). This, too, is an association.

For ordinary reports of suicide, not about celebrities, the same meta-analysis found no association: RR 1.002 (0.997–1.008; 5 studies) per article. The median follow-up here was one day, and an effect for particular types of reporting cannot be ruled out.

Results differed widely between studies, and there were signs of publication bias: studies that find no effect are published less often.

Can stories of coping with a crisis help?

In Austria, a study of 497 news articles from January to June 2005 found that articles about coping with a crisis were linked to fewer suicides, while repeated coverage of the same story and articles repeating myths about suicide were linked to more. This is a study at population level, with no data on individual people, in one country over six months; the authors say such articles may have protective effects (Br J Psychiatry 2010).

A 2022 meta-analysis of individual participant data from randomized trials (8 studies, 2,350 participants) found that stories of hope led to a small drop in suicidal thoughts, a small effect (SMD −0.22, −0.39 to −0.04; 6 studies), only among people vulnerable to suicide (569 of the 633 most vulnerable participants). There was no difference in willingness to seek help (0.14, −0.15 to 0.43; 4 studies). The outcome was thoughts, not deaths, and the effect was small (Lancet Public Health 2022).

In an online randomized trial with 266 adults with recent suicidal thoughts or a suicide attempt, an article by a person who had lived through a suicidal crisis lowered scores on a suicide-risk scale, a small effect (d = −0.18), and did not raise short-term risk. The scale was measured after one week, and deaths were not studied (Br J Psychiatry 2020).

What does the WHO advise journalists?

The WHO's 2023 quick reference guide for media says that widely spread stories about suicide deaths are often followed by a rise in suicides, and that stories of coping with a crisis may lead to a fall. This is the WHO's position.

What the WHO advises doing:

- tell stories of how people cope and of why it matters to seek help;

- say where people can get help;

- give facts about prevention;

- interview the family and friends of the person who died, and people with lived experience of suicide, with care; take special care when reporting the suicide of a celebrity;

- remember that journalists themselves can be affected.

What the WHO advises not doing:

- make the story the top news or repeat it;

- describe the method;

- name the place;

- use sensational headlines;

- romanticise suicide or present it as a solution;

- use photos or video;

- report details of suicide notes;

- reduce it to a single cause.

These points are from the WHO's short guide that summarises the 2023 document.

What happened in Vienna?

Suicides in the Vienna underground rose sharply from 1984 to mid-1987. In the second half of 1987 they fell by 75%, and the lower level held for five years. This is one city and a before-and-after comparison without a control group (Soc Sci Med 1994).

The fall began when the Austrian Association for Suicide Prevention issued guidelines for the media and agreed with them not to report on suicides at all. The timing matches, but this is not proof that the guidelines caused the fall.

Do media guidelines themselves work?

The authors of the 2020 BMJ meta-analysis concluded that responsible-reporting guidelines are the best available intervention at the population level. This is the authors' conclusion; the effect of the guidelines was not measured in that study.

A 2026 meta-analysis of media guidelines (15 studies, 11 countries) found that after guidelines were issued, the media more often followed the advice not to describe the method (RR 1.33, 1.15–1.53; 12 studies). No conclusion about the effect on deaths can be drawn: only three studies looked at it. All the studies were before-and-after comparisons at population level, with no data on individual people (Crisis 2026).