Related Experiment Video
Updated: Jan 2, 2026

Inverse Probability of Treatment Weighting Propensity Score using the Military Health System Data Repository and National Death Index
Published on: January 8, 2020
Suicide Case-Fatality Rates in the United States, 2007 to 2014: A Nationwide Population-Based Study
Andrew Conner1, Deborah Azrael2, Matthew Miller3
1Quinnipiac University, North Haven, Connecticut (A.C.).
Background:
The suicide case-fatality rate (CFR)-the proportion of suicidal acts that are fatal-depends on the distribution of methods used in suicidal acts and the probability of death given a particular method (method-specific CFR).
Objective:
To estimate overall and method-specific suicide CFRs and the distribution of methods used in suicidal acts by demographic characteristics.
Design:
Cross-sectional study.
Setting:
United States, 2007 to 2014.
Participants:
Suicide deaths (n = 309 377 records from the National Vital Statistics System) and nonfatal suicide attempts requiring treatment in an emergency department (ED) (n = 1 791 638 records from the Nationwide Emergency Department Sample) or hospitalization (n = 1 556 871 records from the National [Nationwide] Inpatient Sample) among persons aged 5 years or older.
Measurements:
Rates of suicide deaths and nonfatal suicide attempts, overall and method-specific CFRs, and distribution of methods used, by sex, age group, region, and urbanization.
Results:
Overall, 8.5% of suicidal acts were fatal (14.7% for males vs. 3.3% for females; 3.4% for persons aged 15 to 24 years vs. 35.4% for those aged ≥65 years). Drug poisoning accounted for 59.4% of acts but only 13.5% of deaths; firearms and hanging accounted for only 8.8% of acts but 75.3% of deaths. Firearms were the most lethal method (89.6% of suicidal acts with a firearm resulted in death), followed by drowning (56.4%) and hanging (52.7%). Method-specific CFRs were higher for males and older persons. The distribution of methods varied across demographic groups.
Limitations:
Results are based on suicidal acts resulting in an ED visit, a hospitalization, or death. Consequently, the reported CFRs are larger than they would have been had the data included nonfatal attempts that did not result in an ED visit.
Conclusion:
Variation in overall suicide CFR between sexes and across age groups, regions, and urbanization is largely explained by the distribution of methods used in suicidal acts.
Primary Funding Source:
Joyce Foundation.
Related Concept Videos
Actuarial Approach
Consider the example of a high-risk surgical procedure with significant early-stage mortality. A two-year clinical study is conducted,...
Hazard Rate
Cancer Survival Analysis
Assumptions of Survival Analysis
Life Tables
Survival Curves
The Kaplan-Meier estimator is the most common method for constructing survival curves. This...

