Background
In 2020 in the United States (US), suicide was among the ten leading causes of death among individuals 10–64 years of age [
1]. About 46,000 people died by suicide in the US in 2020, for an age-standardized suicide mortality rate of 13.5 per 100,000 population [
1]. The fact that these numbers include decedents of different types of suicide incidents is rarely acknowledged. In addition to decedents of solitary suicides (a suicide that involves only one decedent), suicide statistics include decedents who died by suicide following homicide, for which the suicide decedent was the perpetrator, as well as decedents who died by suicide together as part of a pact. Despite suicides following homicide and suicide pacts being relatively rare in comparison to solitary suicides (approximately 2% and less than 1% of all suicides, respectively [
2‐
4]), the differentiation of such incidents may have important prevention implications.
It has been found that compared to decedents of solitary suicides, decedents of suicides following homicide are more likely to have been male, non-white, and perpetrators of domestic violence [
4‐
7]. With respect to comparing suicide pacts with solitary suicides, the literature is sparse and outdated, with only two existing studies, both of which use nationwide samples from England and Wales [
2,
8]. Based on these studies, compared to decedents of solitary suicides, decedents of suicide pacts are more likely to have been middle-aged, in a relationship or married, and suffering from a physical illness [
2,
8]. Although there are studies that compare suicides following homicide and suicide pacts to solitary suicides individually, no study to date has compared these two types of suicide incidents to one another.
Given the relative rarity of suicides following homicide and suicide pacts, existing studies are often plagued by small sample sizes. Further, existing studies on such incidents often rely on data sources with limited information (e.g., death certificates). As a result, the characteristics used to describe decedents of suicide following homicide and suicide pacts have been extremely limited in scope. For instance, there is a dearth of research comparing the preceding circumstances, mental health status, and toxicology findings of decedents of solitary suicides, suicides following homicides, and suicide pacts. This represents a significant gap in the literature, as an understanding of such characteristics and the differences between the three types of suicide incidents may provide insight into points of contact where intervention and prevention can occur. Thus, the objective of the current descriptive study was to empirically compare the demographic factors, method of suicide, preceding circumstances, mental health status, and toxicology findings between decedents of solitary suicides, suicides following homicides, and suicide pacts.
Discussion
The current comparative descriptive analysis shows that although there are a few commonalities between solitary suicides, suicides following homicide, and suicide pacts, these three types of suicide incidents represent distinct phenomena. The numerous statistically significant differences found in the various comparisons illustrate this, and these differences have important public health implications. Distinguishing between the decedents of the different suicide incident types provides insight into points of contact where intervention and prevention can occur, which appear to be relatively distinct for each of the suicide incident types. For example, solitary suicides had the highest proportion of decedents with mental health problems and mood disorders, which suggests that mental health professionals and treatment can play an important preventive role in such incidents. In contrast, there was a comparatively lower proportion of decedents of suicides followed by homicide with physical health issues, mental health problems, past suicide attempts, and those who had disclosed their intent. This suggests that prevention efforts within the health care system may not be as effective in preventing this suicide incident type. Suicides following homicide had the highest proportion of decedents with legal and interpersonal relationship problems as well as the highest proportion of incidents involving active methods of suicide. Increased odds of firearm use, an active method of suicide, in suicides following homicides compared to solitary suicides has been previously reported [
4]. Taken together, the legal system may be an area where targeted individual-level prevention efforts could prove to be beneficial for suicides followed by homicide, particularly in regards to domestic issues that escalate to the point of legal involvement [
12]. At the population-level, harsher gun-restriction policies could be a beneficial prevention mechanism.
Suicide pact decedents were on average older and had the highest proportion of physical health problems, as well as opiates present in toxicology screening. This indicates that primary health care providers and health systems may play a crucial role in the prevention of suicide pacts. The high presence of opiates among suicide pact decedents (who were, on average, older than decedents of other suicide types) likely reflects their use of prescription opioids for physical ailments associated with aging such as chronic pain, cancer, etc., as opposed to use of illegal or street-sourced opioids. This underscores a potential increased opportunity for prevention through various contact points within the health care system such as oncologists, palliative care specialists, and/or pain management physicians. It is also interesting to note that despite the high proportion with physical health issues, a lower proportion, compared to solitary suicides, had a mental health problem or mood disorder. This finding could indicate that there is a need for better coordination and collaboration between health care system sectors in order to address the mental health needs of those experiencing physical health problems. There were also a lower proportion of suicide pact decedents with BAC ≥ 0.08 g/dl when compared to the other suicide incident types. This suggests a potential lower level of impulsivity more planning and a longer prevention interval. Finally, suicide pacts had a much lower proportion of decedents who used an active method of suicide compared to the other suicide incident types.
Further, understanding the demographic differences of these three types of suicide incidents will help to identify target groups for prevention efforts. The World Health Organization’s
Live Life approach to suicide prevention suggests “early identify, assess, manage and follow up anyone affected by suicidal behaviours” as an effective evidence-based intervention to prevent suicide [
13]. Characterizing decedents of solitary suicides, suicides following homicide, and suicide pacts, can ultimately inform who is most at-risk and thus, aid in the early identification of individuals likely to engage in suicidal behavior.
The results of the current study are comparable to previous studies comparing solitary suicides to suicides following homicide, as well as solitary suicides to suicide pacts. Aside from the findings of the current literature summarized above, previous studies identified firearms to be more commonly used in suicides following homicide than in solitary suicides [
4,
6,
14], which is reflected in the higher proportion of active methods of suicide in suicides following homicides found here. However, it is worth highlighting that the current study is the first to compare a number of variables across all three suicide incident types (e.g., military status, education, and presence of amphetamines, cocaine, opiates, just to name a few). Further, this is the first study to compare solitary suicides, suicides following homicide, and suicide pacts simultaneously.
With that being said, the current study is not without its limitations. First, despite there being 44 participating states in 2019, not all states are proportionally represented, which also limits our ability to compare suicide types by geographic region. Second, the majority of preceding circumstances and mental health variables are binary variables in the NVDRS. As such, although we can be sure that “yes” means the variable is present, we cannot be sure that “no, not available, and unknown” means the variable is not present. NVDRS data abstractors are limited to the information included in investigative reports; reports that are incomplete, inaccurate, or unavailable may lead to underreporting or misreporting of circumstances and characteristics for some decedents or incidents. This leads to missing data, the extent of which differs by variable, which can be appreciated by the counts provided in the Appendix (Tables A
2-A
5). Third, as per the NVDRS data sharing agreement, some variable categories had to be collapsed in order to supress cells with fewer than ten deaths. In addition to other variables (e.g., method of suicide), this was the case for race/ethnicity. Given that the US is a diverse nation, collapsing race/ethnicity into binary categories may therefore not accurately represent the US population. In the future, when more data becomes available (particularly for suicide pacts), reporting disaggregated data might reveal inequities and disparities in subpopulations that may be obscured by reporting aggregated data. Finally, toxicology findings are only available for decedents for whom a toxicology examination was performed, and the presence of the respective substances were ascertained. Toxicology testing is not supported by Centers for Disease Control and Prevention funding and thus depends greatly on local resources. As such, some states have limited toxicology data due to the cost of testing [
15]. In such states, toxicological data are most often collected only from decedents for whom this information is important for the determination of the cause of death. It is possible that when an active method of suicide is involved the cause of death is more easily identifiable than when a passive method is used; and as such, suicides that involved an active method could have lower rates of toxicological testing, compared to passive methods. In addition, there is a potential racial disparity as certain races may be more likely to have an autopsy performed (for example, the white race may be tested less often than other races [
16]).
This study is intended to lay the foundation for future studies and will hopefully inspire further investigation into the similarities and differences between the three suicide incident types under investigation here. Areas of future research should explore whether the differences noted here vary by age group.
Conclusion
Overall, based on the findings of the current investigation, we can conclude there are notable differences between solitary suicides, suicides following homicide, and suicide pacts with respect to the demographic factors, method of suicide, preceding circumstances, mental health status, and toxicology findings of decedents. This differentiation of suicide incident types is intended to provide insight into the prevention pathways that need to be exploited, some of which are likely untapped.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.