Development of a Computerized Adaptive Test Suicide Scale-The CAT-SS.

Area of Science:

• Psychiatry
• Psychometrics
• Mental Health

Background:

• Current suicide risk screening is inefficient and lacks precision.
• Existing methods focus narrowly on suicide-specific items, neglecting related symptom domains.
• There is a need for a more balanced and comprehensive assessment of suicidal symptomatology.

Purpose of the Study:

• To describe the early development of a computerized adaptive testing (CAT) suicide scale.
• To psychometrically harmonize suicide, depression, and anxiety symptom domains for a complete spectrum of suicidal symptomatology.
• To pave the way for the final validation stage of the CAT suicide scale.

Main Methods:

• Collected data from 789 psychiatric outpatients (70% female) with a 47% major depressive disorder rate.
• Developed an item bank of 1,008 items covering depression, anxiety, mania, and suicide.
• Used a bifactor model and simulation to create a computerized adaptive test (CAT-SS) from 308 subjects' responses.

Main Results:

• Identified 111 items extending suicidality assessment to include related depression and anxiety domains.
• A mean of 10 items achieved high correlation (0.96) with the 111-item scale, offering 5-point precision.
• Preliminary validation showed a 52-fold increase in suicidal ideation likelihood across the CAT-SS range.

Conclusions:

• The CAT-SS accurately measures the latent suicide dimension using a mean of 10 items in approximately 2 minutes.
• This scale offers a more precise and efficient method for suicide risk assessment.
• Further validation against clinician assessments and prediction of suicidal behavior is ongoing.