JAMA Psy: Machine learning identifies risk factors for suicide attempts

Depression, a common mental disorder, is now the second-largest killer of cancer in humans, with an estimated 350 million people worldwide suffering from depression.
The World Health Organization estimates that depression will be the world's largest burden of disease by 2030.
According to data published in The Lancet in 2018, the global incidence of depression is about 6 per cent, while the lifetime risk is 15-18 per cent, meaning that one in five people experience depression at some point in their lives.
half of them live in Southeast Asia and the Western Pacific, including India and China.
, suicides due to depression have become more frequent in recent years.
there is a close link between psychosocial stress and suicidal behavior, but this link can not occur independently, but through some kind of intermediary role, after all, not all people who experience stress will commit suicide.
, it is possible to understand suicide in greater depth only if the personal traits of the suicidal person are analyzed and the risk factors that are suicidal are combined.
machine learning is a multi-disciplinary interdisciplinary subject, involving probability theory, statistics, approximation theory, convex analysis, algorithmic complexity theory and other disciplines.
specializes in how computers simulate or implement human learning behavior to acquire new knowledge or skills and reorize existing knowledge structures to continuously improve their own performance.
has also been widely used in medicine in recent years.
We've all heard that big data promises to change the medical world through a widely available collection of electronic health files and high-volume data streams that can be used as a source of data, from insurance reimbursements to personal genomics and biosensors registration systems.
From the application of risk scoring to guide the use of anticoagulant therapy (CHADS2) and cholesterol-lowering drugs (ASCVD) to risk stratified (APACHE) in patients in intensive care units, data-driven clinical prediction has become the norm in medical practice.
, combined with modern machine learning, clinical data sources allow us to rapidly generate predictive models for thousands of similar clinical problems.
from early warning systems for sepsis to super-human imaging diagnosis, the potential scope of application of these methods is considerable.
machine learning be used to identify risk factors for astray suicide in the general population? The answer is yes.
team from Columbia University in the U.S. looked into the issue and the results were published in the latest journal JAMA Psyhiatry.
the study data are from two surveys conducted by the National Epidemiological Survey of Alcohol and Related Diseases (NESARC) (2001-2002) (2004-2005).
a face-to-face longitudinal survey of a nationally representative sample of the U.S. non-institutional civilian population aged 18 and over.
response rate in the two surveys was 70.2 per cent, resulting in 34 653-wave interviews.
used cross-validation to train a balanced random forest to develop a suicide attempt risk model.
model predictions are used to evaluate the performance of the model, including area, sensitivity, and specificity under the receiver operating curve.
of the 34,653 participants, 20,089 were women (52.1 per cent) and 222 (0.6 per cent) reported suicide attempts within three years between interviews.
Using the survey questions in the first interview, the suicide attempt risk model produced a cross-validated subject operator characteristic curve with an area of 0.857, with a sensitivity of 85.3% and a specificity of 73.3% at the optimized threshold.
model identifies 1.8 percent of the U.S. population as at a 10 percent or higher risk of suicide attempts.
The most important risk factors were three questions about past suicidal thoughts or behaviors, three items in 12 short-form health surveys, i.e. frustration with emotional problems, a low sense of accomplishment in the work done, a younger age, a lower level of education, and a recent financial crisis.
, the study looked at more than 2,500 survey questions and identified several well-known risk factors for suicide attempts, such as past suicidal behavior and mindth.
also identified new risks, including mental disorders and socio-economic disadvantages.
these results may help guide future clinical assessments and develop new suicide risk scales.
reference: Factor of Suicide Attempt Risk Factors in a National US Survey Using Machine Learning. JAMA Psychiatry. Published online January 6, 2021. doi:10.1001/jamapsychiatry.2020.4165MedSci Original Source: MedSci Original Copyright Notice: All noted on this website "Source: Mays Medicine" or "Source: MedSci Original" text, images and audio and video materials, copyrights are owned by Metz Medical, without authorization, no media, website or individual may reproduce, authorized to reproduce with the words "Source: Mets Medicine".
all reprinted articles on this website are for the purpose of transmitting more information and clearly indicate the source and author, and media or individuals who do not wish to be reproduced may contact us and we will delete them immediately.
at the same time reproduced content does not represent the position of this site.
leave a message here