From the Journals

Machine learning identifies childhood characteristics that predict bipolar disorder


 

FROM THE JOURNAL OF PSYCHIATRIC RESEARCH

Opening doors to personalized care

Martin Gignac, MD, chief of psychiatry at Montreal Children’s Hospital and associate professor at McGill University, Montreal, said the present study offers further support for the existence of pediatric-onset bipolar disorder, which “remains controversial” despite “solid evidence.”

Dr. Martin Gignac, chief of psychiatry at Montreal Children’s Hospital and McGill University, Montreal

Dr. Martin Gignac

“I’m impressed that we have 10-year-long longitudinal follow-up studies that corroborate the importance of this disorder, and show strong predictors of who is at risk,” Dr. Gignac said in an interview. “Clinicians treating a pediatric population should be aware that some of those children with mental health problems might have severe mental health problems, and you have to have the appropriate tools to screen them.”

Advanced tools like the one developed by Dr. Uchida and colleagues should lead to more personalized care, he said.

“We’re going to be able to define what your individual risk is, and maybe most importantly, what you can do to prevent the development of certain disorders,” Dr. Gignac said. “Are there any risks that are dynamic in nature, and that we can act upon? Exposure to stress, for example.”

While more work is needed to bring machine learning into daily psychiatric practice, Dr. Gignac concluded on an optimistic note.

“These instruments should translate from research into clinical practice in order to make difference for the patients we care for,” he said. “This is the type of hope that I hold – that it’s going to be applicable in clinical practice, hopefully, in the near future.”

The investigators disclosed relationships with InCarda, Baylis Medical, Johnson & Johnson, and others. Dr. Gignac disclosed no relevant competing interests.

Pages

Recommended Reading

Telemental health linked with improvements in key outcomes
MDedge Psychiatry
Brief Psychiatric Rating Scale succeeds as transdiagnostic measure
MDedge Psychiatry
Preparing patients with serious mental illness for extreme HEAT
MDedge Psychiatry
Lithium for bipolar disorder: Which patients will respond?
MDedge Psychiatry
Lithium, valproate, and suicide risk: Analysis of 98,831 cases
MDedge Psychiatry
The hunt for N-acetylcysteine: Medicine or dietary supplement?
MDedge Psychiatry
Clinical psychoeconomics: Accounting for money matters in psychiatric assessment and treatment
MDedge Psychiatry
‘Disturbing’ lack of follow-up care after psychiatric crises
MDedge Psychiatry
Bipolar risk and parental age: What’s the relationship?
MDedge Psychiatry
Lamotrigine for bipolar depression?
MDedge Psychiatry