Sometimes social media seems like a mindless exchange of duck-face pictures, LOLs and emoticons. But what if those feeds contain something more? What if some social media exchanges hide an encrypted cry for help — a darker truth, unknown even to the author?
That hypothesis intrigued Michael Birnbaum, MD, a psychiatrist at the Zucker Hillside Hospital, who studies social media use and language patterns among young people with psychosis. His goal is to identify and predict mental illness in patients between 15 and 35 years old to enable intervention before psychosis occurs.
Social media work in behavioral health is similar to that being done in other fields, Dr. Birnbaum said. “It bridges the gap between computer science and behavioral health, bringing mental health care to the digital age.”
Dr. Birnbaum explained that using technology to improve clinical outcomes in behavioral health has lagged behind other areas. While the study of social media-based linguistic and behavioral analysis is in its infancy for psychiatric disorders, Dr. Birnbaum’s studies — published in Early Intervention in Psychiatry — show it may be a reliable predictor of psychotic symptoms.
“Psychiatry is all about early intervention, outreach and engagement,” he said. “The research suggests that by studying language patterns and social media usage, and monitoring changes in those areas, it may be possible to determine the likelihood of a psychosis relapse before it can be clinically detected. That is a huge differentiator that would benefit patients because it would enable an intervention before a psychiatric episode even occurred.”
Through research projects at Zucker Hillside, Dr. Birnbaum is analyzing how young people with psychosis (schizophrenia and schizoaffective disorder) use the web — specifically social media. Collaboration with a linguistic analysis team in Texas and James Pennebaker, a psychologist at the University of Texas, who studies the psychological meaning of word usage and computerized text analysis through a scientific “linguistic inquiry word count” method, can determine the linguistic style of people who suffer from depression and other mental illness.
“The pilot data suggests the computer algorithms may be able to identify individuals who have psychosis and who are in the midst of a psychotic event,” Dr. Birnbaum said. “As a possible predictor of the onset of illness or worsening of symptoms, it has the potential to positively affect among of these patients.” Dr. Birnbaum is applying for additional grants to expand and conduct further research in this area. While it is still early in the research phase, based on the results, he said he is optimistic a breakthrough for patients and their families is on the horizon. Dr. Birnbaum studies how young people with psychosis use the mobile web.