January 07, 2021
2 min read
Save

Nearly 2% of U.S. population may be at significant risk for suicide attempt

You've successfully added to your alerts. You will receive an email when new content is published.

Click Here to Manage Email Alerts

We were unable to process your request. Please try again later. If you continue to have this issue please contact customerservice@slackinc.com.

Researchers have used an algorithmic approach to confirm well-known risk factors for suicide attempt and identify new ones, according to study results published in JAMA Psychiatry.

“More than one-third of people making nonfatal suicide attempts do not receive mental health treatment,” Ángel García de la Garza, BA, of the department of biostatistics at Columbia University, told Healio Psychiatry. “Given this, we wanted to extend our understanding of suicide attempt risk factors beyond high-risk clinical populations to the general adult population. The National Epidemiologic Survey on Alcohol and Related Conditions (NESARC) data set provides an extensive evaluation of substance use, psychiatric disorders and symptoms not routinely available in electronic health records, giving us a unique opportunity to do so.”

infographic with quote about suicide risk assessment

García de la Garza and colleagues used a data-driven machine learning approach to analyze data of 34,653 noninstitutionalized individuals aged 18 years or older in the United States. Analysis included more than 2,500 questions from the large, nationally representative NESARC, with data available from its first wave between 2001 to 2002 and its second wave between 2004 to 2005. The researchers developed a suicide attempt risk model using a balanced random forest, which was trained using cross-validation. Further, they used out-of-fold model prediction to evaluate model performance, including the area under the receiver operator curve, sensitivity and specificity. Estimates were representative of the U.S. civilian population according to the 2000 census. Main outcomes and measures included attempted suicide in the 3 years between first and second wave interviews.

Results showed 222 (0.6%) participants self-reported attempted suicide between the waves. Survey questions measured at the first wave showed the suicide attempt risk model produced a cross-validated area under the receiver operator characteristic curve of 0.857. At an optimized threshold, the sensitivity was 85.3% (95% CI, 79.8-89.7) and the specificity 73.3% (95% CI, 72.8-73.8). According to the model, 1.8% of the U.S. population was at a 10% or greater risk for suicide attempt. The most significant risk factors included the following:

  • Three questions about previous suicidal ideation or behavior;
  • three items from the 12-Item Short Form Health Survey, namely feeling downhearted, doing activities less carefully or accomplishing less because of emotional problems;
  • younger age;
  • lower educational achievement; and
  • recent financial crisis.

“We hope our results provide a novel avenue for future suicide risk assessment,” García de la Garza said. “We believe that the inclusion of the three questions from the 12-Item Short-Form Health Survey in future suicide risk scales could improve the assessment of suicide attempt in clinical practice.”