Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

Related Experiment Videos

Suicide prediction revisited

A D Pokorny1

  • 1Baylor College of Medicine.

Suicide & Life-Threatening Behavior
|January 1, 1993
PubMed
Summary
This summary is machine-generated.

Reanalyzing suicide risk in psychiatric inpatients using logistic regression proved ineffective. The method identified too few actual suicides and too many false positives, limiting its clinical utility for suicide prevention.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by
Same author

Comment: validity of self-report in alcoholism research.

Alcoholism, clinical and experimental research·1989
Same author

Alcohol use and psychosocial outcome of two preventive classroom programs with seventh and tenth graders.

Journal of drug education·1988
Same author

Stress, coping, family conflict, and adolescent alcohol use.

Journal of behavioral medicine·1987
Same author

Alcohol use by adolescents in disrupted families.

Alcoholism, clinical and experimental research·1986
Same author

Borderline personality traits in psychiatric inpatients.

Psychological reports·1986
Same author

Selected behavioral features of patients with borderline personality traits.

Suicide & life-threatening behavior·1986
Same journal

A Randomized Controlled Trial With an Internal Pilot of a Co-Designed App Targeting Upstream Suicide-Related Risk and Protective Factors Among International Students.

Suicide & life-threatening behavior·2026
Same journal

Predicting Next-Day Passive Suicidal Ideation in At-Risk Youth.

Suicide & life-threatening behavior·2026
Same journal

Emotional Responses to Social Media and Non-Suicidal Self-Injury Among Adolescents.

Suicide & life-threatening behavior·2026
Same journal

Prevalence and Timing of Suicidal Ideation, Preparation, and Attempts Among Mass Shooting Survivors.

Suicide & life-threatening behavior·2026
Same journal

Suicidality of Indigenous Teenagers as an Anomic Adaptation to Frustration: An Empirical Test (Correlates of Indigenous Teenage Suicidality).

Suicide & life-threatening behavior·2026
Same journal

Feasibility, Acceptability, and Impact of the Couples Crisis Response Plan (CCRP) to Reduce Post-Discharge Suicide Risk: A Randomized Controlled Trial.

Suicide & life-threatening behavior·2026
See all related articles

Area of Science:

  • Psychiatry
  • Epidemiology
  • Biostatistics

Background:

  • Accurate prediction of suicide risk among psychiatric inpatients is crucial for effective clinical intervention.
  • Previous analyses of a 1983 cohort yielded limited utility for identifying patients at high risk of suicide.
  • Logistic regression is a statistical method often considered for binary outcome prediction.

Purpose of the Study:

  • To reanalyze suicide risk in a cohort of psychiatric inpatients using logistic regression.
  • To assess the predictive accuracy and clinical utility of logistic regression for suicide identification.
  • To explore methods for improving the performance of logistic regression models in this context.

Main Methods:

  • Reanalysis of data from a 1983 prospective study of 4800 psychiatric inpatients.

Related Experiment Videos

  • Application of logistic regression analysis to predict suicide outcomes.
  • Simulated analyses involving manipulation of base rates and hypothetical predictive tests.

    • Main Results:

    • Logistic regression demonstrated poor performance, identifying too few subsequent suicides (low sensitivity).
    • The analysis generated an excessive number of false positives (low specificity), limiting practical application.
    • Artificial adjustments, such as increasing base rates or adding hypothetical tests, showed improvement but did not achieve perfect prediction.

    Conclusions:

    • Logistic regression, as applied, is not sufficiently accurate for identifying suicide risk in psychiatric inpatients.
    • The inherent challenges in predicting rare events like suicide limit the utility of standard statistical models.
    • Further research is needed to develop more effective predictive tools for suicide prevention in clinical settings.