A 10% increase in suicides — nearly 2,000 additional deaths — was recorded in the United States in the four months after actor and comedian Robin Williams took his own life in 2014, according to research published Wednesday in the journal PLOS ONE.

The “celebrity-suicide effect” — in which copycat suicides follow that of someone famous — has been documented in previous research.

“This is the first study to examine the consequences of a celebrity suicide in the digital era,” said David S. Fink, lead author of the new study and a post-doctoral researcher in epidemiology at the Mailman School of Public Health at Columbia University.

Williams, 63, was found dead in his Tiburon, California, home August 11, 2014, of what investigators suspected was a suicide by hanging.

Fink began his research by forecasting the number of suicides that could be expected to occur between August and December 2014. He and his colleagues analyzed monthly suicide rates in the United States, based on data gathered from January 1999 through December 2015 by the US Centers for Disease Control and Prevention.

Based on Fink’s analysis, 16,849 suicides would be expected from August to December 2014.

Yet 18,690 suicides were reported in the four months after Williams’ death: an additional 1,841, or a 9.85% increase, compared with the expected number. Both females and males as well as all age groups showed more-than-expected suicides.

However, the highest number of unexpected suicides was observed among men (1,398 excess suicides) and people between the ages of 30 and 44 (577 excess suicides).

The researchers also found a greater-than-expected number of suffocation suicides — which includes hanging — in the months after Williams’ death: a 32% increase in the method used by the comedian himself. By comparison, other methods of suicide rose by just 3%.

Media reports — some of which described how Williams hanged himself — might have provided the “capability necessary for a high-risk segment of the US population, middle-aged men in despair, to move from suicidal ideation to attempt,” Fink said.

“We found both a rapid increase in suicides in August 2014, and specifically suffocation suicides, that paralleled the time and method of Williams’ death,” Fink said. He also found, by conducting a news trend search on a Bloomberg terminal, “a dramatic increase in news media reports on suicides and Robin Williams during this same period.”

Though he cannot say for sure that Williams’ death inspired copycat suicides, the significant rise in suicides coupled with frequent media reports about the actor’s death suggests a connection, he said.

“This study supports much of what we already know about the influence our environment has on our behaviors in general and suicide in particular,” Fink said. Policy-makers and researchers need “to better understand the role that traditional and social media have on suicidal behaviors to understand how to better mitigate these deaths.”

However, there’s one problem, said John Ayers, a computational epidemiologist who works as an adjunct associate professor at San Diego State University.

“The problem with this study is: ‘So what?’ ” said Ayers, who was not involved in the research.

Ayers said Fink’s new research is “rigorous science; it’s a well thought-out study; it asks an engaging question and provides a very precise set of conclusions.”

“But there’s no urgency there,” he said.

“Robin Williams died in 2014,” Ayers noted, and all the people who were vulnerable to following the actor by committing a copycat suicide have done so.

Just as the common cold is contagious, so too is suicide.

If you witness self-destructive behavior within your family, within your peer group or through the media, you may be at greater risk for suicidal behavior yourself, according to the National Institute of Mental Health. “Suicide contagion” affects some people, including those who face a chronic medical condition or who are depressed, more than others.

“What we need to do is extend this research beyond living in the past to the present,” Ayers said. “And try to develop strategies where we can have these types of insights while they’re happening so we can respond.”

Celebrity influence occurs in all areas of life, he said. But when it comes to suicide, celebrity copycat effects are a concern for public health experts, including the World Health Organization, which established media guidelines for reporting a high-profile death.

“The WHO standards on how to discuss suicide in the media, one, they’re well-known, and two, they’re based on decades of research,” Ayers said. He explained that the guidelines include “don’t mention the modality of suicide, don’t dwell on the suicide act or speculate on the reasons why, and paint suicide as a rare event that should be avoided.” And media should never sensationalize or normalize suicide or present it as a solution to problems, he said.

In Williams’ case, Fink and his colleagues found that traditional media headlines deviated from the WHO guidelines. For example, headlines of some media reports speculated on the cause of his death while others divulged details of his suicide.

“We don’t know much about how these events happened and what type of message is spread that may be violating these standards when it comes to how we report on suicide,” Ayers said. “They may ultimately be the culprit for increases in suicide ideation or actual suicide attempts or suicide itself.”

Just as traditional media reports about Williams’ death and suicide drastically increased in the weeks after his death, so did posts on social media sites.

“Reddit has been a tremendous success in terms of how to provide a forum for people to discuss suicide safely,” Ayers said. “The suicide forums on Reddit are carefully curated by employees at Reddit who are experts in suicide prevention and awareness.”

Reddit monitors posted content to provide a safe place for people to discuss suicide. But if a discussion moves into the realm of “tacit encouragement” of suicide, Ayers says, Reddit steps in and curtails the conversation.

Fink points out that many people now receive the news of a celebrity suicide via social media. And so Williams’ death stands “in contrast to another high-profile US entertainment star suicide: Kurt Cobain, lead singer of the enormously popular rock band Nirvana, in 1994,” Fink said.

Cobain’s death had minimal impact on the subsequent suicide rate in his hometown of Seattle, though suicide crisis calls increased in the days after his death. One study indicated that traditional media reports restricted details of his death and provided consistent messages regarding suicide prevention.

This careful reporting “may have played a pivotal role in preventing subsequent suicides,” Fink said. By comparison, social media posts may not align with WHO guidelines and so may act “as a new and emerging risk factor” after the suicide of a celebrity.

Ayers said that in certain jurisdictions in Canada, “when you talk about suicide on certain social media forums, they actually inform the Mounties, right? And the police come and intervene when they know you’re on the edge of committing suicide.”

“You have a potential opportunity here,” he said. The classical finding that the celebrity effect can cause copycat suicides can be used to consider how to find people online when they are at risk and in need of help and how to intervene, he said.

In his own recent study, Ayers tracked internet searches for “suicide” after Netflix’s release of “13 Reasons Why,” a fiction show that explores a teen’s suicide. Believing that the program did not follow the WHO suicide guidelines, he called for Netflix to remove it until the producers got the messaging right.

Ayers hopes to create “a rapid response ecosystem for public health” in which social media can “be a pathway by which we listen,” he said. “If we spent more time listening to the public and anticipating their needs, we could intervene when it’s happening and respond.

“Places like Reddit are emblematic of the kinds of successes we can have where people go to seek help and get help,” Ayers said. “On all other social media channels, what type of help are people getting? Unfortunately, we don’t know.”