Can you explain the five classifications you built as part of the study?
Crisis: These were youth who had died shortly after some sort of interpersonal or scholastic crisis had occurred in their lives that was readily apparent to the CDC.
These youth did not have indications of prior mental health concerns by and large. Clinicians are familiar with evaluating people who might have been severely affected after a major event. For example, admitted inpatients after their first major relationship breakup.
Disclosing: Somewhere along the way, these individuals told somebody that they had had suicidal thoughts or intentions and that could be detected. These youth may or may not have been in treatment previously.
Identified: Typically, individuals who die by suicide tend to be male, but this class was majority female. They tended to be in mental health treatment and were more likely to have died by asphyxiation or poisoning.
Hidden: These individuals were not disclosing, were not clearly in crisis, and did not have elevated rates of substances in their postmortem toxicology reports. We called these kids hidden, and they are an important group of individuals to try to understand better because they are a very sizable minority. This class was 21% of individuals in the study. But also consider that about half of individuals across classes had had no prior mental health contacts or disclosures in the CDC data. This is a group that may be a new beachhead to reduce suicide rates in children because we currently do not have a clear way to identify, reach and treat them.
Surveillance: There’s one final group that’s called surveillance.
We used data from the CDC’s National Violent Death Reporting System or NVDRS which happens to have limitations in how representative it is. States markedly differ in the degree to which they report and when they started participating in this database. So this classification represents individuals who do not have enough data to clearly resolve their archetype. This class takes its name from the type of database the NVDRS is, “surveillance data”, which is well understood to have this limitation.
I should note that Colorado has been an early participant in the NVDRS, and this database has taken the monitoring and prevention of suicide very seriously. And so we have a bit more information about our youth deaths in Colorado, for example.
Could there be overlap between these classifications?
Absolutely. Humans tend to think in archetypes, so it does help us as a field to know each classification is representative of the story of many individuals. The practice of medicine can sometimes be variable-centric, too. Like focusing on your hemoglobin A1C level, outside of anything else. Doctors constantly go back and forth between treating the variable and treating the person. This study is more on the person side, less on the variable side.
There might be people in the identified class who did not disclose, for example, but they had died by asphyxia, had been in mental health treatment with all these other indicators, just didn't happen to disclose. There's going to be some variations on specific variables.
Now these four classifications are what the math said were the most clearly resolved groupings. That doesn't mean there aren't subclassifications and more nuance. It does help suggest that we can't have a one-size-fits-all approach to suicide prevention.
And did you and the team come to a conclusion on what some of those current gaps would be, especially as it comes to detection?
It's very clear from the data that there are different phenotypes, and we need to weigh our policy options going forward in responding accordingly. That said, it starts with awareness of resources and then the continuation and expansion of universal screening efforts.
Awareness of resources
- Individuals in crisis are helped by programs such as Safe2Tell in Colorado or 988. They provide on-demand touchpoints during a crisis that can provide real-time support. Letting parents, caregivers and kids know this is an option should be a priority.
Safe2Tell Colorado is an anonymous line to report danger to someone’s safety. Reports can be made online or by calling 1-877-542-7233.
988 Colorado is a 24/7 support line for emotional, mental health or substance use help.
Both resources are free and can be done anonymously.
En Español
Continuation and expansion of universal screening
- Detection of suicide is an emerging trend in pediatrics, and the mandate to do suicide screening in hospitals has come about only in the last decade. Expanding universal screening to primary care visits will clearly reach more kids because they are mandated to see their pediatricians. There are validated tools, like the ASQ (Ask Suicide-Screening Questions) toolbox, to help this effort.
- There are other systems, such as education, that have profound reach, but will likely need support to permit screening.
What is impulsivity and why is just studying adults insufficient when it comes to suicide?
We should talk about development. Impulsivity is a major driving risk factor for youth suicide – especially for individuals in the very early pre-teen group, it's a leading risk factor.
That's clearly tied to development. We're not fully well-developed in terms of emotion regulation and general impulsivity until our mid-20s and later. Youth challenges also include navigating early relationships and identity formation. That leads to some unique pressure points around hopelessness, worries about the future, the intensity of feelings of those new relationships. And finally, unfortunately, there is definitely an aspect of contagion effects within youth – where there are multiple suicides across a peer/classmate group.
As providers, we also have to think about the connections between impulsivity and lethality – as that’s a key interaction to suicide risk. Firearms, jumping, asphyxia all have dire, irreversible consequences. That’s why it’s important to study and discuss things like firearm storage safety.
What's one key takeaway from this research and broader work in youth mental health that you'd want people to know?
Suicide risk is detectable and we can prevent it.
We now have access to a large degree of nuanced data and other studies that are making a huge impact on identifying who is at risk. For instance, some types of sleep disruption can increase one’s odds of having suicidal thoughts and behaviors. This increase in risk lasts for years, a more extended period than we have appreciated in the past.
There's a substantial opportunity at this point to really respond and get better at detecting and preventing suicide in kids.
Key Points:
- About half of youth who die by suicide have no prior mental health histories or contacts with the mental health system.
- Youth suicidal behavior, compared to adults, is often more likely to be impulsive, creating challenges in reaching adolescents and teens prior to attempts.
- A new typology of youth suicides – including four primary categories – is instructive in showing where the gaps in the healthcare system are.
- The first steps in reducing youth suicide are awareness of resources and a continuation and expansion of universal screening for suicide risks in other healthcare settings.
.jpg)
