Research, Practice, and Data Informed Investigations of Child and Youth Suicide – Day One
SHELLI AVENEVOLI: Good morning, everyone, and welcome. Welcome to the National Institute of Mental Health. We very much appreciate your taking time out of your busy schedules to spend the next two days with us to discuss and collaborate on addressing the important aims of this meeting.
As many of you may know, the NIMH is a lead federal agency for mental health research. We fund a broad portfolio that covers the full spectrum of mental disorders as well as the full translational continuum of science. Suicide prevention is and has been one of our top research priorities. Today’s workshop focused on child and youth suicide is a critical and timely component of these research priorities. We are all well aware that suicide rates for both children and adolescents have been increasing. It is also the case that suicidal ideation and attempts accounts for an ever-increasing number of emergency department and in-patient visits at children’s hospitals nationwide.
We also know that there are considerable racial and ethnic disparities in rates of suicide and that has been a disproportional increase in rates of suicide among black youth. Collectively, these trends reflect the larger mental health crisis our youth are experiencing today.
These increasing rates of suicide and concerns about the mental health impacts of the COVID-19 pandemic have brought increasing attention to the area of youth mental health and at the highest levels of our government. For example, in 2019, the Congressional Black Caucus released a report Ring the Alarm: The Crisis of Black Youth Suicide in America.
At the end of 2021, we also saw the release of the Surgeon General’s advisory on protecting youth mental health. And most recently, a White House report on mental health research priorities included suicide prevention in youth mental health as top areas of research and urged federal agencies to work together to address these problems.
These issues we face are complex and require and demand our collaboration. It is in that spirit that we bring you together today.
Not long ago, we began a collaboration with other NIH institutes and other federal agencies to develop and host a series of workshops and roundtables that brought together a variety of perspectives on the topic of suicide risk among youth.
Today’s workshop builds on these prior discussions and focuses on the death investigation process and capitalizes on a collaborative planning effort between NIMH, the CDC, and the Maternal and Child Health Bureau. Many of you bring to the table a variety of diverse perspectives, those of the death investigation community, the child death review and violent death data collection community and the child and youth suicide research community. We very much look forward to hearing your perspective and to begin to outline how we might work together to resolve our challenges and to work together to determine, assess, and ultimately mitigate suicide death in children and youth.
Thank you for having me today and I now will turn it over to Dr. Joshua Gordon.
JOSHUA A. GORDON: Hello. I am Dr. Joshua Gordon, director of the National Institute of Mental Health. Thank you for your interest in this NIMH workshop on research, practice, and data-informed investigations of child and youth suicide. Suicide is complicated and tragic, especially when it strikes the young. But it is also often preventable. Therefore, suicide prevention research is a top priority for NIMH.
This workshop builds upon past NIMH workshops and roundtables, devoted to the topic of child and adolescent suicide. In those meetings, questions were frequently raised about the challenges involved in suicide death determinations in children. It is with these challenges in mind that we have asked this distinguished group to come together to discuss the practice of manner of death investigations in suspected suicides in children and the types of research that could inform these practices. It is our hope that your work here will lead to science-to-practice and practice-to-science approaches to this tragic yet growing problem.
NIMH thanks our sister agencies, the Center for Disease Control and the Health Resources and Services Administration, who work with us to develop this workshop agenda. We particularly appreciate the introductions to representatives of the medical examiner, coroner, and death investigation communities.
We further thank you all for participating, knowing that you have invested considerable time and effort in preparing for and attending this workshop. Throughout the workshop, we hope that you will identify opportunities to collaborate with others to advance our understanding of child suicide and to incorporate that knowledge into actions that can lead to enhanced suicide investigation and prevention practices. Thank you again.
Agenda Item: Overview of Child and Youth Suicide Statistics and Challenges with Data Collection and Surveillance Activities
MARGARET WARNER: Good morning. I am Margaret Warner. I am a senior epidemiologist and a director of the Collaborating Office for Medical Examiner and Coroners at the National Center for Health Statistics, which is one of the Centers for Disease Control. Today, I am going to be giving an overview of child and youth suicide statistics and a little bit about data collection and surveillance activities to get us started for the next two days.
We will first talk a little bit about the practice of death investigation, and we have medical examiners and coroners and forensic pathologists and death investigators in the room who will give us their introduction on how these suicides among young persons are investigated. We will talk a little bit about the data that is being collected and some of the ways we use those data and how youth suicide statistics are collected and how they inform research.
And then the researchers will tell us some about the state of the knowledge about youth suicides and youth suicide ideation in children and youth. They will be informing the practice of death investigation. And we are hoping that after the end of the meeting, we will have a lot of collaboration on all sides of this circle of getting out information to inform prevention programs and the meeting is two days and really looking forward to it.
I am starting out talking about some statistics, which I am going to start presenting now. We know that suicide is a leading cause of death among persons 5- to 24-year-olds in 2021. This chart shows the five leading causes of death among persons 5 to 24. The suicide ranking is in red. Among persons 10 to 14, suicide is actually the second leading cause of death. In that age group, we have a reported 598 deaths in 2021. Among persons 15 to 19 and 20 to 24, suicide is the third leading cause of death with 2343 suicides reported among 15- to 19-year-olds and 3877 among persons 20 to 24 years old.
This information that I just showed you is based on information from the National Vital Statistics System. It is the Vital Statistics Mortality Data. Sometimes you will see this data referred to as Death Certificate Data or Vital Statistics Mortality Data or sometimes it is referred to as the Platform that the data is extracted from so WONDER or WISQARS. But all the information – those are just different names of the data. They are all the same data and they are based on information from death certificates filed in the United States and the territories.
The data include death demographic information such as age, sex, race, and geographic location as well as occupation and industry beginning in the year 2020. It also includes causes of death information that are coded to the International Classification of Diseases, Tenth Revision. And the states provide data to the CDC through a cooperative agreement.
This is a US standard death certification. In fact, most states have the majority if not all the deaths reported in an electronic form now. But the death certificate is something that we can use as a way to share what is collected. We have demographic information, which is completed by the funeral director as well as cause-of-death information, which is – for natural causes is most often completed by the attending physician. Sometimes it is a nurse practitioner or physician’s assistant, and this varies a little bit by state. But for sudden and unexpected deaths, including suicides, that is completed by a medical examiner or coroner or Justice of the Peace.
I have blown up here the cause of death section because it is such an important component to the reporting of suicides. We have in the death certificate a causal sequence leading to death as well as contributing conditions. And the manner of death is a checkbox, and this is the most critical piece for determining or reporting whether the death was a suicide. You can see here that there are six boxes, one of which is suicide. When the medical examiner or coroner were filling out this death certificate if they believe this death to be suicide related, they will check this box. Also, included is how the injury occurred. And the suicide would be considered an injury so this box would be completed.
Just a little bit about what medical examiners and coroners do. The medical examiners and coroners and the death investigators during the next session will talk a lot about how they do this. This is just a very basic summary. A death is reported to the medical examiner and coroner and then they investigate the death. They look a little bit about the circumstances and history surrounding the death. They may attend the death scene, talk to witnesses. They may review medical reports. In the case of suicide, they may try to find out about mental history and also legal history. They may do an examination of the body, including a physical examination, including sometimes an autopsy. They may do ancillary tests. Oftentimes with suicide, they may check toxicology to see whether any drugs or alcohol were on board and genetics and other tests may be done. Then they take all that information into consideration and they will determine the cause and manner of death and fill out the death certificate.
These practices do vary across the country and we could talk about that at a later date about the investigation and how it is done. And the medical examiners and coroners will speak to this as well. But a quality investigation will include most of these pieces for a suicide-related death.
The manner of death, as I mentioned, is that checkbox. Here is just a little bit more about that checkbox with some definitions. These were developed by the National Association of Medical Examiners in collaboration with CDC and it is in our instruction manual on how to fill out the death certificates or what we call our handbook on death registration and death reporting. The handbook is in brown and then the manner of death guidelines is the white document. I have the resources listed here if you would like to look those up. They are freely available on the internet if you just Google it. But the link is listed there as well.
Suicide results from an injury or poisoning as a result of an intentional, self-inflicted act committed to do self-harm or cause the death of one’s self. The medical examiners and coroners again will talk about this in greater detail.
This slide is a little bit complicated, but I did want to put it up because it talks a little bit about how the information gets from the death certificate all the way to the endpoint of those data systems I mentioned. As I mentioned when I talked about the death certificate, the medical examiner and coroners or the physician fill out the medical component of the death certificate and then demographic information is filled out by the funeral home or funeral director based on what their next of kin reports. That information is all the way over to the left-hand side of the screen.
In the middle portion, they fill out a death certificate and that is sent to the electronic death registration system, which in turn goes to the National Center for Health Statistics, which is the other box in the middle of the screen. And then we code and process this information at NCHS and that goes back to the state at EDRS as well as it goes into these national systems.
I want to highlight up at the top of the data of that slide. Some of this information goes into state and local public health and public safety data systems, including child death review as well as the National Violent Death Reporting System. And each state has a state violent death reporting system as well as a state child death review.
But information from both the medical examiner and coroners as well as the EDRS, the death certificate information, feeds into these systems. I mention that because we have representatives from both child death review and violent death reporting system here in the room and they will talk about the systems as well as a couple of other of these state-based or local-based public health and public safety data systems. These folks can collect a lot more information on the death, a lot more detailed information and at a later time. Death certification takes place pretty quickly after a death for a variety of different reasons. With additional information that they get, they have more information to inform prevention. Those systems also feed up to a national system, that is, both child death reviews and violent death reporting system feed up to other national systems and that is in the upper right-hand corner.
But I just point out to you that the vital statistics death certificate information as well as the medical examiner and coroner systems play a critical role in feeding both those systems information to get more detail as well as they make violent death inclusion information from law enforcement and child death review and a few other sources in certain states. There is a lot of information that they have in addition to the death certificate. But there is definitely a relationship between the death investigation done by the medical examiner and coroners and those systems as well as the coded causes of death from the vital statistics and mortality data electronic death registration systems.
NCHS, where I work, I mentioned it many times already. We do a lot of promoting the consistency and uniformity in the data. We have many different activities, including model vital statistics laws. We have the manuals that I mentioned. I mentioned the program or the handbook that describes how to fill out a death certificate. We have some automated software to help with coding the causes of death.
I mentioned this because sometimes questions about the quality of the data and we do have some programs to help with that and welcome questions about these systems if you are interested in these quality checks.
Now, I am going to get a little bit more into the statistics, just a bit of detail about the number of suicides. This is the number of suicides by single year of age for persons ages 10 to 17. This is from 2018 to 2021. I have taken here an average annual number of suicides in order to have the data meet our standards of reliability. But you can see from the ages of 9 to the ages of 17 in this 4-year period where I have taken the average number of suicides. From among 9-year-olds, the average number for this 4-year period is 12. That does meet our standards of reliability for reporting. You can see that the number does increase all the way to the age of 17 where there is an average of 574.7 deaths reported. But it does increase by age.
One of the reasons I wanted to show you this slide is because the systems do allow reporting of children under 10. There is some misunderstanding about what we allowed to be reported versus what we do report. We tend to not report deaths among children less than 10 because of the issues with the reliability of the numbers. I can share with you more of the statistical details if you want. But the bottom line is there are very few. It is difficult to calculate statistics. And I would say that is a fortunate thing.
This slide shows the same basic information so suicides among this – now, I am starting at 10- to 17-year-olds over time. The first darker blue bar is among the years 2010 to 2013. The lighter blue is 2014 to 2017. And the last green bar is 2018 to 2021. Again, these are not averages. This is the rate. This is a population-based number. We have taken the number of deaths in this period and divided it by the population in these age groups during those years. Again, you can see the increase.
Among children 5 to 9 – in order to show you a little bit among children 5 to 9, I have combined all the years from 2002 to 2021 and taken an average annual number, again, taking the number of deaths reported and dividing it by that time period. You can see that there are very few but it does seem to be increasing in the last four-year period.
Over this total period, there were 129 suicides among 5- to 9-year-olds from 2002 to 2021. Among those 5- to 9-year-olds, I have changed the time period here from 2001 to 2021. There are 136 suicides among 5- to 9-year-olds. Seventy-eight percent of them were among males. Eighty-eight percent of them were suicide-related causes, primarily hanging. There are other ways – suicide or asphyxiation. But primarily, these are hangings.
By race, there is variability. The white, non-Hispanic, crude rate is .02 per 100,000 so quite low. Among the black, non-Hispanic, the crude rate is .08 per 100,000. The Hispanic crude rate is .02 per 100,000.
I wanted to talk a little bit about the quality of the demographic and personal information on the death certificate because I know it is going to come up quite a bit during this presentation. And again, we have representatives from the medical examiner and coroner community. But most of the demographic information on the death certificate does come from funeral directors.
We do have some linkage studies to evaluate the quality of the race and Hispanic origin on the death certificate. And the results show misclassification of these data with 3 percent for Hispanic origin and 3 percent misclassification for Asian and Pacific Islanders as well as a 34 percent for American Indian and Alaska Natives. We do plan to publish an update of this study in late 2023 to also include additional data items.
The report highlights the importance of improving what we call the entry points of information. We are working more with funeral directors. That is not a focus of this study but we do have a program to begin working more closely with funeral directors in a more systematic way than we have in the past.
In terms of improving data from death certificates, we do believe that it is very important to improve the data at the source. And that is one of the purposes of the next two days is to work together on death investigation methods, bringing together the practitioners doing the death investigation, talking a little bit about the data and then we are really focusing on some of the research. How can these three groups work together to improve the data at its source?
As I mentioned, we have electronic and interactive resources for training cause-of-death certifiers and funeral directors. We have some tools for funeral directors, and we have interoperability of the data systems, the EMR and the medical examiner systems.
I have already mentioned our new program to focus on funeral director programs and as well as – I will talk just briefly about the Collaborating Office for Medical Examiners and Coroners, which we have recently developed at NCHS.
This is a slide showing our website. We are not all about a website, but a lot of information is on our website. We have information for medical examiners and coroners on death certification as well as a lot of these different systems to try to bring together all the resources we have at CDC to support their work.
Again, I am really excited about the next two days to talk about how we can do the science-to-service and service-to-science approach where we are working on the practice of death investigation, how data is collected, and at what points in time it is collected and the different sources of information and the research that can inform the death investigation practice.
I would like to really highlight again the slide where I was talking about the important role medical examiners and coroners as the starting point of death investigations of potential suicides among young children and young adults. That information is fed into the vital statistic system and is used for a lot of those different follow-on data-driven activities such as violent death reporting and child death review. They really rely on the information from the medical examiners and coroners as a starting point to investigate the death, learn as much as they can so that they can both inform the families but also learn what they need to know to prevent the deaths in the future.
If anybody has any questions, I have up here my name and number. I am happy to share or talk with you. Thank you.
BRENDAN WEINTRAUB: Hello everybody. Welcome back. We are going to do a quick poll before we move on to our next session. For folks who have not used this before, we are using something called Mentimeter. I will walk you all through it in just one second. What I would like everyone to do is go to www.menti.com and then use the code 31733. We have just a quick question just to get folks involved today. What stakeholder group do you most align with? We have researcher, medical examiner, coroner, fatality review team member, a family survivor, or other. This is available for folks in the room or discussants outside of the room and then of course everybody else who is watching. So far, it looks like we have a few others. A lot of researchers here, some medical examiners. We will give it another minute. It looks like we have a lot of researchers here, a few others.
But I wanted to just thank everyone for attending today. I wanted to see how we can actually use the information that we are learning today and it looks like we have a lot of researchers in the room. It is going to be a great opportunity to hear more about data.
I am going to pass it over to Marcus Nashelsky and he is going to introduce the next session.
Agenda Item: Overview of Current Approaches to Child and Youth Death Investigation When Manner Is Potentially Suicide
MARCUS NASHELSKY: Thank you, Brendan. Good morning. What a pleasure after a year of virtual meetings to meet many of you in person. It is good to be here.
I am a forensic pathologist and a medical examiner in Iowa. My professional home is the Department of Pathology at the University of Iowa Hospitals and Clinics.
Our goal this morning for this group and for the many virtual attendees is to offer a crash course in medicolegal death investigation. My hunch is that many of you here and virtually have had focused exposures and perhaps many focused exposures with a medicolegal death investigation system. But the overview may not be part of your experience even for the experienced researchers here.
With four sessions over about 90 minutes, we will describe for you how good medicolegal death investigation works in the United States. We will start with a general description of optimal medicolegal death investigations by Dr. Joyce deJong of Western Michigan University. Dr. deJong is the chief medical examiner of many counties in Southwest Michigan and she is the current president of the National Association of Medical Examiners.
Following Dr. deJong, Lauri McGivern will describe for us the role and work of medicolegal death investigators. By this, we mean the individuals who work in high-functioning death investigation systems who take initial reports of death, go to scenes, and collect often a mountain of data about the decedent whose death is being investigated. Lauri is from the Office of the Chief Medical Examiner of Vermont. She is the long-time director of all field investigators, medicolegal death investigators in the State of Vermont. Lauri is the past president of the American Board of Legal Death Investigators.
Our third presenter will be Dr. Kathryn Pinneri from Montgomery County, Texas. Dr. Pinneri will further focus the conversation on difficulty investigations, primarily centered on the difficult and uncommon, we think, cases of suicide in the young.
Finally, Dr. Jan Gorniak will be our discussant. Dr. Gorniak is a forensic pathologist in Clark County, Nevada, working with the coronary medical examiner office based in Las Vegas. Dr. Gorniak has experience in leadership positions in Atlanta, Georgia, Washington, DC, and Columbus, Ohio and now in Las Vegas.
We have asked Dr. Gorniak to distill the content of the first three presentations and we hope to bring a personal and very experience-based comment about what you have heard for the first hour and how she sees the difficulty case type of death investigation of the young where the manner of death is presumed to be suicide. Thank you all for joining us, especially out there in virtual land.
I will now hand it off to Dr. Joyce deJong.
JOYCE DEJONG: Good morning and thank you so much for this. I am Joyce deJong. They kind of cut off where I was from because it is a real long title. Western Michigan University Homer Stryker M.D. School of Medicine. I think we got through Homer or something but I do not say it all very often. But we are in Kalamazoo, Michigan. I am going to talk to you briefly about optimal medicolegal death investigation systems, recognizing that not all of them are optimal.
There are a number of questions. What do we do when there has been a reported death? You want to know who is this person, so identification. That is kind of an issue. But these are the basic questions. Why did you die? What caused this? We want to know when the person died. There is a manner of death and you are going to hear a bit more about that, what is the manner. And then how did this happen? If you were injured, when did this happen and how did this happen? Those are the general questions that we have.
Now, if you think about that and, Margy, you showed this slide earlier, a portion of a death certificate. These are not questions that we just made up. They were directed to us by the death certificate. I would like to get an A on a test that I take. I like to get all the questions right. Right down who is this person and all of those times as best as we can. These are the questions that we get asked.
Oftentimes in our profession, we come under a lot of criticism for things like manner of death a lot of times like how do you know it was a suicide? Why did you call it a homicide? Why did you do that? Who are you? Who are you to make that determination? That should be up to the jury. And this was something actually the death certificate and manner of death is an invention in the United States. It was made to provide information to public health. This was not for the legal system. This was to help the nosologist classify that. Because if you had a gunshot wound to the head, that is important to know. But did the person kill themselves? Were they shot by another? Did the gun drop and go off? That happens on occasion, pretty rarely, but it happens. It was to be able to classify that really. I look at it like where is our problem and where do we need to focus our efforts.
Looking at that then, how do we do this? We start by – we get a history and go out to the scene. We send our investigators out, gather this information. We look at other information. We talk to law enforcement. We look at medical records. We look at mental health records, EMS reports, policy reports. We are examining the body, that is part of it. We will do additional testing. There might be some follow-up testing that we need. And then we issue reports. We issue documents with this on. It might be the death certificate, the autopsy report, investigative reports. We will have toxicology reports. That is kind of the system that we have. It is a little bit like very much – if you think about it, we are physicians as medical examiners, as forensic pathologists. And the clinical physician really does something very similar. You go to see your doctor. They are going to ask you why are you here and what happened. We can try asking our patients that. They do not respond. If they do, it is a really disturbing moment. You have to ask other people. You have to know about this information. Your physician will look at your medical history, social history, family history. They will order tests, order x-rays just like this, additional testing, history. They might do different things. They will issue a document in the medical record, diagnosis. They will do treatments and surgeries. But it is a very parallel kind of process for us.
What do we do now? Cause of death. That is the big question. One of the major questions. What caused your death? These are just a couple of stab wounds but there are hundreds of different causes. It could be a gunshot wound. If we are talking suicide, it could be a gunshot wound. It could be hanging. There are so many others. It could be drug related. But there are hundreds of different causes of death that we can identify. And sometimes it is more difficult. Sometimes the actual autopsy and examination of body will appear that everything is normal. Then you have to really look into the circumstances of the death and looking at your toxicology results and looking at how was this person found and what else happened. You need to take in all of that information. That would be your cause of death.
And then you are looking at the manner of death. And like you mentioned, we are limited really to five. This is a gunshot wound to the head. I am not going to show too many disturbing photos. But I will say for the pathologists in the room, this came in as a suicide. This is not a contact range gunshot wound, which most of the suicides are. You are going to ask questions. I can tell you right now that if that is the cause of death, it is not natural. We can eliminate that. But it is an accident, a suicide? Is it a homicide? Or in some cases, we have inadequate information to tell. How are we going to make that determination on what the manner of death is?
We are going to do all of those things like we said. We are going to look at the history. We are going to look at the circumstances. But when we are looking more like is this is a suicide, we are going to ask some additional questions as well. Were they alone at the time that this happened? If this happened and there was another person in the room, that is a little more disconcerting especially given it is not contact range. But who found this person? Was the house secure? Where is the gun? Who owns that gun? When was it purchased? Was there any declaration of intent? Were there any departure notes there? Were there previous threats? What were the physical findings of the autopsy? There are on occasion times where we say this gunshot wound was probably from four or five feet away. And unless there is an elaborate device set up for this, it probably is not. But we are looking at all this information and asking a lot of questions about each death.
These are our questions. Somebody has reported a death. We want to know this. There are some other questions too. But these are the basic ones that we are looking at. When you think about medicolegal death investigation systems, a lot of times people are like it is the morgue. It is the morgue and its autopsies. It is simple. How hard can this be? But it is actually pretty complicated. It is a complicated system and it may appear straightforward but it is a complex process. Our systems are complicated. How do we do this?
If you look at the process, a death gets reported. You have to have somebody to report to, not just law enforcement. You have to know which deaths need to be reported and who are they reported to. Most of our reports come from law enforcement or from hospitals or from health institutions. On occasion, we will have a family member call and say I just found someone – how they jumped to calling a medical examiner’s office is always a little bit interesting. Nobody is faulted for that. We do the investigation by the medical examiner or the coroner. They will respond to the scene. They ideally will respond directly to the scene. They will talk to the family. They will talk to anybody else around there and they will gather this information, what we are going to also ask for other data. We are going to ask for their medical records, for mental health documents.
Those are oftentimes challenging to obtain. If we really need them, we can typically try and work with the family if they are interested in that. But there are a lot of protections understandably about that information because there is oftentimes a lot of information about individuals other than the family in there who are still alive. We will do additional interviews.
An autopsy will be done. A lot of times the information that we need is not necessarily in the autopsy and examination of the body. We will do all of this. And then, like I said, we will issue these reports.
We are really talking about the optimal setting here, the optimal situation. There are a couple of programs. The IACME, which is the International Association of Coroners and Medical Examiners, and the National Association of Medical Examiners. They both have accreditation processes. The accreditation is for the system. It is for the office and the system. It is not directed at the individual practitioner. Is your system adequate? Do you have the appropriate resources to do quality medicolegal death investigation? Not every office in this country is accredited. I think we are at over a half of the population of the US being served by accredited offices but there are some major systems that are not. That does not mean they are doing a bad job. But there are these programs.
For NAME accreditation, I know there are probably about 400 or so. Does anyone know the exact number of lists on the accreditation checklist. Everything from are your outlets appropriate and grounded to. Do you have a toxicologist who is board certified and are your pathologists board certified? There is just a long list of things that you have to do for this. It is for the system.
The National Association of Medical Examiners also has autopsy performance standards. These are really to guide the professional, the forensic pathologist in the practice of their art of medicine. These are the things you should do. It covers what kinds of – which deaths should you investigate and which deaths should require an autopsy? What things should you do in certain autopsies if it is a stab wound, if it is somebody who is unidentified? What are the specific things that you should do in each one of those cases to guide the individual?
Both the accreditation and the NAME standards are at a minimum level. And oftentimes offices – they should at least meet them and the individual practitioner should meet the standards. But it is okay to exceed them and it is actually encouraged. But at a minimum, this is what should occur.
If you look at NAME accreditation, it covers all of these different areas. There are a lot of generalities. What should happen and what resources do your investigators need? What should be occurring in the morgue? When I look at things like morgue operations here, it does give you the basics. For example, we now in our office have a CT scanner. That is not even mentioned in any of these programs. Now that I have one, I think every office needs one. It is incredibly valuable for gathering information and for being able to even demonstrate that information and do a 3D reconstruction. You can actually show people – fractured patterns and things like that. But this is the minimal and these are the areas that we talk about in the accreditation. This is what is required for optimal death investigation systems.
I will just talk a little bit about some of these. Do you have an appropriate facility? Do you have a facility? Is it secure and is your data secure? Are things safe? Are things safe for the people that are there? Is it adequately maintained? And really importantly, I probably think one of the biggest ones here is a quality assurance program. Are you looking? Does everybody just get to work in a silo and turn out what they want or is there some mechanism here to say I am going to look over your work and make sure that you are covering everything appropriately?
The quality assurance, like I mentioned there. You want to have consistent standards. You want to have continuous improvement. You want to have your people to be adequately trained. I think one of the really important aspects of this as well is when I look at this, are you being objective? Are you being objective? I have seen a number of cases that – and sometimes it is painful to be objective. But I know I have been involved in – in Michigan, death investigation varies by state. It is the state’s statutes that define it.
In Michigan, we are medical examiners but to be a medical examiner, the requirement is you have to be a physician licensed to practice medicine in the State of Michigan. In other states to be a medical examiner, you have to be a board-certified forensic pathologist. In the states with coroners – I have made no disparaging remarks about coroners. They are partners in our death investigation. We try to speak with a unified voice. But in many states to be a coroner, you need to be at a certain age and have a GED. There is a wide variability on what that is. That is pretty uncommon.
I want to point out then that the system needs to be independent and not under the influence of other entities. Sometimes that influence really is pressure from families even saying this is not a suicide. On occasion, I have had them saying this is a suicide but there is a lot of pressure. And to maintain that objectivity can be pretty challenging.
A quick one. Autopsy begins at the scene. This is a 47-year-old woman, driver of a minivan. When we talk a little bit about manner of death, this occurred at 1 in the morning. This is a widow. She broke up with her boyfriend about a year ago. She had some depression. No prior suicide attempts. But it was pretty clear that she died of her injuries. I kind of like the idea of if we can change our manner of death to this like if it is a clearcut homicide, you are there. If it is I cannot really tell if it was an accident or a suicide. We could just put it right there. We can do things like that.
In her case, up against a tree, 1 o’clock in the morning. She had her 14-year-old daughter in the car with her. That seems a little weird. She is going down the road and she just hits a tree. But it is 1 in the morning. Maybe she fell asleep. Then you find the cell phone. She texted her boyfriend, if I cannot have you – if you will not have me, I may as well just wrap my car around a tree. Suddenly we go from accident or suicide too. We are over here. But that kind of information is really important.
We have standardized protocols. We have standardized protocols in our system just for suicides. And we even have standardized protocols for youth assessments. That is an important part. I do not think they are perfect. I think we can improve them. I think information from a collaboration like this would help us to gather better information, better data to share with you about what is going on. Collaboration and coordination are really important.
We have board-certified ABMDI death investigators and your pathologists should be board certified by the American Board of Pathology. Our senior investigators, and Lauri are going to talk more about that, they are boarded by the American Board of Medicolegal Death Investigators. We have the American Board of Pathology.
I will make one final comment then and that is that resourcing is really difficult because people do not want to pay for this. Actually, just at a dinner last night, there were a few of us talking. We can have all of these things. If we had one request, one ask, that would be that every office had mental health professionals in the office to help with the families because we go out to families, especially in a suicide or any kind of some traumatic event.
If we had social workers, other mental health professionals out there or to take those calls to facilitate them, it is not a requirement but that would probably be the biggest asset and probably prevent significant number of suicides if they had those resources because we see them coming back, other family members a couple of years later. That is an optimal death investigation system.
Are you coming back up Marcus or am I going to turn it over to Lauri McGivern? I am just going to let Lauri go.
LAURI MCGIVERN: Thank you for having me here and speaking on such an important subject. Disclaimer from the first slide. I am not Dr. Lauri McGivern. I am a board-certified medicolegal death investigator. I just wanted to clear that up before we move forward.
Today, I am going to talk to you about what we, as death investigators, do to collect the information that you use for your research and other things and what data resources we use, how the data is actually used, some of our challenges in the field when we are collecting the data, and some considerations for future – for accurate and timely data collection.
As you have heard, standards and accreditations. Ideally, your medicolegal death investigator is certified by the American Board of Medicolegal Death Investigators. They are trained and they follow standards that are set forth by the national committees and also, they work in a medical examiner or coroner or federal authority office that is accredited through the National Association of Medical Examiners or the International Association of Coroners and Medical Examiners.
Our data collection consists of a whole lot of people. It is not just the medicolegal death investigator that collects data on the scene. We have a death investigation team. We do work independently but we are all gathering information for the same purpose. We want to know what the cause and manner of death is.
Oftentimes when we are responding to scenes, we are responding with law enforcement and there is some oversight also with a state or a district attorney. The medicolegal death investigator is trying to gather as much information as they can to help the medical examiner or coroner determine cause and manner of death.
Law enforcement is also interested in cause and manner of death but their focus really, as they are gathering information, is to make sure that there is no criminal aspect to the investigation. As I mentioned, the state or district attorney will have oversight of all of that and make determinations if there is a criminal aspect whether charges are brought or there is prosecution.
Our data collection starts right at the time of dispatch. When I receive a call, I am already gathering basic information about the scene, basic demographics, who the person is, what their name is, date of birth, the location, and things like that. As I am responding to the scene, I may get more calls where I gather more information.
As I arrive at the scene, I am going to meet with law enforcement and typically on the scene before I am. I am going to gather that information from them as well, which will lead us a little bit more into what that incident was, how it evolved. Also, we may have EMS on scene if they are still there and they can help us with some of the information gathering as well.
If we have witnesses, bystanders, or family on scene, we will interview them at the time. We also will reach out to primary care providers, ask about medical history, ask about any mental health history. We will do medical record reviews as well so we will request medical records, which will be reviewed. In some cases, we will review social media sites. We may look at the individual’s Facebook or Twitter account and see if there are any indications of self-harm. If by law we are able to and this depends on the jurisdiction, we may look at computers that are on site, tablets, and phones as well.
Our data collection in any death no matter what the manner is, we start really from general to specific. We have to gather information that we know is going to help the medical examiner/coroner with the cause and manner of death but also help with filling out the death certificate. We know we need basic demographics. We need information about the death, the location, if there was an injury where the injury was, how the injury occurred.
Again, continuing with that general information, the past medical history. It can be challenging sometimes to get some of these records, as was previously stated, especially the mental health providers or some of those notes. But if we can, we will try to have conversations or reach out to them to see what the mental health was of that individual, look at problem lists. We will look at medication lists. And a lot of the times on the scene, we have medications as well. We will actually review the medications. We will count the medications and we do that along with law enforcement.
We also will start to gather information about next of kin who may not be on the scene. We want to make sure that we have that contact so that we can reach out to next of kin and again gather some more information.
Continue with demographics. Again, this information is used for the death certificate. As we start to become more specific in the questions that we are asking, we want to know about that living situation for that individual. Did they live alone? Did they live with a partner? Were they in a congregate setting?
When we suspect that the manner of death is suicide, we are going to ask even more specific questions. Was there a declaration of intent? Again, looking at social media feeds. Sometimes there are statements that are made by an individual that may be a declaration of intent. We may have verbal or written notes, a suicide note. Although it is not as common as people think. It is 20 percent or less that we will see on the scenes. And we are seeing less and less written notes. Oftentimes there may be a note or some indication in a phone or a text to someone or on the computer. Or was it demonstrated? Are there things laid out? Everything is planned. The bank account numbers are here with the passwords. The will is here. Are there some other demonstrated things on scene that may indicate that this is a suicide?
We want to know about previous threats and attempts. We are going to ask about the date. We are going to look into what type of method was used for that previous attempt. We will also ask about behavioral changes. Did you notice the individual? Was there a behavioral change? We will go back 24, 48, 72 hours and ask friends or family that may have been with this person during that time.
Diving into the mental health history if we have access to it, we want to know if there has been a diagnosis, any inpatient hospitalizations or outpatient therapy and medications that that individual may have been taking.
We will also ask about family mental health history. When I am sitting with a family, you can start to see, as I go through the rest of these questions, these become really personal for these families. They are grieving. You are dealing with grieving families. There may be some guilt associated as well with these families. And now I am starting to ask them, is there any history of any mental illness in your family. Can you talk to me about that? Any prior suicides? This becomes very personal.
And then when we talk about the guilt and family dynamics, I am asking – any recent arguments. Yes, we just had an argument. He went outside and shot himself. That is the reality of the work that we do. It is really challenging because these families are already torn apart and we are coming in as strangers, basically sitting down and trying to gather this information.
This is a short list of circumstances but some of the common ones that we do see. Financial problems. A lot of the time there has been some legal problems. Maybe they just got pulled over for a DUI or there is some possible arrest, intimate partner and relationship problems, sometimes school problems. We will ask about those circumstances. Sexual orientation and gender identity. We will ask this if at the time that is appropriate. And then with females, we do want to know about any recent pregnancies, any miscarriages, abortions, anything like that.
And then once we are working on a case that involves a youth, again, we will start to be more specific in our questioning. We want to know about their experience in school. Have they been having any problems in school? Any issues with their grades at all? Any bullying and if it did involve bullying, what type of bullying was it? How did that occur?
Again, continuing with relationship stressors, arguments with family/friends/classmates that they have been with and interaction with law enforcement. I keep bringing this up because it does sometimes seem like a common thing that we see. There has been some interaction with law enforcement.
Online or social media conflict. We are seeing more and more of that as these kids are online quite a bit. We will try to ask that question and even see if there are friends or other people that would be open to us talking to them.
What are the dynamics of the family? Again, this becomes really personal for families as you are sitting with them. Have you had issues? Is this a single family? Are the children going between parents? What does that dynamic look like? Again, very personal questions. I am asking these within an hour or an hour and a half of them finding out that their loved one has died. You can imagine how challenging that is to get the answers.
How is this data used? For medical examiners and coroners, like I mentioned, we are gathering as much as information for our pathologists or the medical examiner to help them determine the cause and manner of death and then ultimately to complete the death certificate. The death certificate is important that that information is accurate and timely. It goes up through the National Center of Health Statistics and is reported out. We want to make sure that the information that is being reported out is accurate.
It is also used – I just have a couple listed here in the national databases, which I am sure all of you are familiar with, the National Violent Death Reporting System, NVDRS, and the State Unintentional Death Overdose Reporting System, SUDORS. Those would be manners of death that are undetermined. As was mentioned earlier, it may be there is not enough information to determine that this was an accidental overdose versus suicidal.
Also, the state death review teams use this information. Child fatality review teams. We are starting to see more and more jurisdictions that have suicide death review teams.
What are the challenges in data collection? Sometimes lack of information on the scene. I may be out in the middle of the woods, recovering a body from a tree that has been hanging and not have family or have other people that are going to give me that information. I could be on the side of the interstate with somebody who had just jumped in front of a motor vehicle. These scenes are not as stable as people think about. We are not just sitting there talking to the families person-to-person. Oftentimes, we may have to seek them out whether they are home afterwards, try to reach them by phone, which is not optimal. It is best to have an in-person interview if you can with that individual. But sometimes the information is just not available at the time. To be honest with you if I spent three or four hours on a scene trying to do a body recovery, I am off to my next case. I am off to the drug overdose. I am off to the motor vehicle accident. I do not have time sometimes to go back and ask all of these questions.
Also, the families can be very emotional and grieving. I think in general, we work very well with families. And we are able to get the information that we can at the time and it really takes a lot of patience. I learned many years ago from a great mentor, the act of being silent. I think it is so important to be able to ask a question, give time for people to answer it and to give periods of silence.
We can sometimes have difficulty obtaining medical and mental health records. It depends on the jurisdiction, the statutes within that jurisdiction. I can tell you within my jurisdiction, we are within the health department, which is great. Our medical examiner’s office sits there.
I do not have issues obtaining the information. I have issues sharing it. We have HIPAA – they are covered entity. We have issues sharing information. That is an area that I think needs a little more work because we can obtain but not always share.
And then lack of time and resources. As I mentioned, medical examiners and coroners’ offices are really busy, a majority of them. We are moving from case to case to case. It is hard to have that follow up or to do those secondary interviews. And sometimes it is not even good for the families to be going back later on and asking some of these questions.
What should we consider in the future? I do believe that death investigators at the time if they have the family there, should collect that information at that time. You have that one chance. You may not get that chance again. Collect as much information as you can, being very sensitive and considerate as you are doing that.
But we do have to prioritize our data. Again, when I am prioritizing, I am gathering as much information for that forensic pathologist so that they can make a determination of cause and manner of death.
I may not be asking the question about the family’s financial status or the individual’s financial status. Did they have state assistance, things that I know we are all interested in for public health purposes?
What are some other things we can do? Case managers or social workers within medical examiner offices or within law enforcement. In my state, a couple of years ago our state police, which is our largest law enforcement agency in the state, hired some social workers. They were informed. They have worked with – they are trauma informed and they have worked with grieving families. That is what you need, people who understand how to approach these families.
When they are on my scenes, it is incredibly helpful. They connect with the family at the time, the same time I am connecting with the family. They are already starting to offer resources and to help this family navigate through the worst day of their life. What do we do next? How do we get a funeral home? How do we get the death certificate? How do we get a medical examiner’s report? This is really the next piece that we need to do to make our offices serve the families and to serve the families much better than we are currently doing. We do not have time to do this and we do not have the expertise. That is the reality of it. We need people to come in and help and to, as was mentioned earlier, prevent further deaths by reaching out to these families really early.
You can use public health nurses. We use that in our states. We use them for infant and child death. I work with a maternal child health group. If there is an infant/child death, those public health nurses reach out to the families.
What do we need? We need additional funding and resources and opportunities, medical examiner, coroners. It is very limited. We have been trying to get a social worker in our office for a very long time. I am currently pulling from three different grants to get maybe a half-time social worker. Very challenging. We need more opportunities to move forward.
That is all I have.
BRENDAN WEINTRAUB: Next up is Dr. Pinneri.
KATHRYN PINNERI: I just want to say thank you for inviting me for this presentation. It is a topic that is near and dear to my heart. I hope that you have gained some understanding about the process that we, as forensic pathologists, go to. It is a snap decision. We just do not say oh it is a suicide and walk away. Although it may seem like that to surviving family members. We really do put a lot of time and effort in these deaths. I am going to talk about some additional challenges that we have as forensic pathologists and medical examiners in determining these manners of death particularly in the young.
I do not have any financial interests or disclosures. Someday I hope to have one but not now.
Manner of death, as you have heard, is a classification. It is really just based on the circumstances in which the death occurred. It is not a legal term. It really was designed for public health statistics.
We have the five options that Dr. deJong described to you: natural, accident, suicide, homicide, and undetermined. And this is the definition of these types of death kind of the way that we approach it. Natural death is something that occurs from within that person’s own body. An accident is a result of something originating outside of that body causing an injury. A suicide is a self-inflicted act intended to cause harm that results in death. A homicide is a volitional act of a person that directly causes the death of another person regardless of intent. It is not a legal term. And then the undetermined. There really is not enough information or we have competing causes of death and we really just cannot pick. Instead of locking something in, we go with the undetermined.
Again, as you heard from Lauri and from Joyce, we use a combination of the autopsy findings and the circumstances. We have to have a good scene investigation from our medicolegal death investigators. But we also sometimes have to use law enforcement information, date of person died at the home – or, I’m sorry was transported to the hospital and then we cannot get back in the home. Law enforcement might have to assist us with that.
We also do review a lot of medical records. It can highlight areas of stress or areas of mental health disorders that they are being treated for. Maybe they are not taking their medications because a lot of times we get – I have gotten huge boxes full of medications and they are all full. They are not taking them because they do not like how they make them feel or they just do not think they are working or maybe they are working. They get better and they stop taking them and then they get worse again. Medications, medical history – very helpful for us in our manner of death determination.
Let us talk a little bit about intent. I put up a couple of definitions here from dictionary.com and Miriam-Webster.com. There is no such thing as a real dictionary anymore. It is all online. Something that is planned, proposed or intended. The act or fact of intending, as to do something. Clearly formulated or planned intention. We have no way of looking at a decedent and what their intent was. We can look at the injuries and we can make assumptions based on those injuries, but a contact gunshot wound to the head is a contact gunshot wound to the head. Without the circumstances and without that scene investigation, we cannot determine the manner of death for those instances. We have to look at everything and keep in mind the possibility of intent.
It is impossible to determine a person’s state of mind based on an autopsy. We really cannot do it. I wish that we had these special powers to see things like schizophrenia and depression in the brains when we are looking at them but they are physiologic processes and we cannot see them unfortunately. We really cannot do anything with state of mind based on an autopsy alone.
I will give you a little personal thing. I am a suicide survivor. My sister killed herself. It has been several years now. It does not matter that they had plans the next day. They had plans the next day, but they also had a plan to kill themselves too. It is really a huge struggle for us when you see these things and I hear these – they planned on this and they planned on that. I know and I understand but they still killed themselves. We have to take everything into consideration and not just look at one particular thing. Plans for the future do not necessarily mean that there were no plans to kill themselves. We have to all keep that in mind.
The injury appearance though can be very helpful for us from an autopsy standpoint because if it is not a contact gunshot wound, I am going to be stepping back a little bit from a self-inflicted injury. It does not mean it could not be. Maybe they are like do I really want to do that. Well, then they still shot themselves and they still did do that. But it would make me consider if there was someone else in the room. Maybe I need to take along the homicide line of manner. We really have to keep an open mind. And I think most of us do.
Again, most of this information comes from the scene and subsequent death investigation. I will not go into the scene investigation too much other than the photographs are very important. We get one chance to do this right. We really want to try to do as much as we can.
Suicide notes, as Lauri said – we are seeing less and less of them. I think kids today are learning to type before they are to write. We are not going to see handwritten suicide notes anymore. They are going to be test messages. They are going to be Facebook or Snapchat or Discord post. They are going to be something along those lines from these younger generations.
We need to see what the location of the weapon is, any presence of drugs or drug paraphernalia. Just because someone is intoxicated, it does not mean that they intended to kill themselves. It is no difference in my opinion – this is my opinion alone. If you get drunk and you kill yourself, it is really not any difference if you get drunk and you get behind the wheel and you get into a car wreck. Maybe you did not mean to get into a car wreck but you still got into that car wreck.
Any nearby electronic devices or journals. Always helpful. Electronic journaling is a big thing now. Blog posts, stuff like that. Electronic devices are huge in our world now. As Lauri mentioned, we may not have the jurisdiction to look through them. Law enforcement might but they might need a search warrant and that is going to come with approval from the parent or the next of kin potentially.
The suicide information out there is growing and growing. There are literally step-by-step instructions and just knowing if someone visited that site or participated in one of those groups is a very important thing for us to know when we are looking at the manner of death determinations.
These electronic devices though are so important for the family members. We really want to give it back to them. They have photos and notes and really it is a lifeline back to the person who died. We really want to give them back to them as soon as possible but we also hope the families understand how critical that information is for us. As sudden and unexpected as it may be for the family members, that information, that phone might provide healing and help to them to know they were really struggling and were hiding that from me.
Again, another use of electronic devices is to look for any cyberbullying or blackmailing, which can be a trigger for an otherwise completely non-depressed person whether it is a child, teen, or young adult.
The good news is there is legal action that can be taken if we are aware of this. They can charge the cyberbullying individual in some jurisdictions. A lot of it is coming from overseas though. However, the local law enforcement can reach out to those jurisdictions if they can track them down through computer forensics.
The social media posts and websites, are very important. Children and teens tend to be more open about their thoughts and feelings with online peers. The last few years with COVID and the isolation and the online communication with others. These are not just online people. These are their friends. These are their almost family. These people are peers that they are giving information to. It is not just an online friend or something to blow off. These are very important people to these kids. They might share more things with them.
We are seeing a lot of these social media challenges. I hate these. Benadryl challenge that just killed a person. The choking game challenge on TikTok just came back unfortunately. And then there is a lot of sexual experimentation talk. It is a lot freer to talk about sexual things when you are not face to face with someone. Kids are starting at a much earlier age with these things. The exposure has really increased at a younger age. The note section may have personal journaling and also making passwords for their devices. We have already gone over most of what we are going to see.
We really want to try to identify in this younger population a triggering event because that is what pushes these kids to do this impulsive act. This is where we can make efforts towards prevention. Because if we can intervene on these triggering events, maybe we can take that anxiety and depression and deescalate it down into something where they realize I can get through this. I can do this.
The depression and substance abuse part of this are important because a lot of individuals are experiencing things like depression. Substance abuse is pretty rampant, especially in our world. In our child fatality review team, we always want to know were they being treated for their depression or their mental health problems. Did they have the insurance to seek that treatment or to get those medications? As Lauri said, we cannot always ask those questions. But if we do get medications, we can see if they were taking them appropriately.
School records are very important in this group too because some kids do not tell their parents everything that is going on in school. The schools do reach out to families when there are disciplinary problems but not necessarily behavioral or bullying type problems. It is important to reach out to the school if you are really puzzled about a case to see if they have any interactions between the counselors and that child.
Again, the child’s friends are an important part of this. Again, it is a tricky situation when you are talking to potentially other minors. This is going to be something that would have to go through this law enforcement based on the guidelines proposed by that jurisdiction. I have had several of these in my career where you have a high achiever who gets a B and it is the end of the world. They are not going to get into Harvard. They are not going to get into this. It is a disgrace and they just cannot handle it. I wish I could tell them it is really not important. I could not even tell you what my grades were. I absolutely got B’s. You can still go on and perform but these grades – they just become such an important part of these teenagers who want to go off to do amazing things and they do not see a way out of recovering from a B.
We take all of these things and put it together with autopsy findings. If I see the presence of cutting marks or scars, that is an important thing. That is not normal behavior. That indicates some sort of stress or depression or need to feel something besides mental things. It is a physical pain they are looking for to distract from their mental pain.
Evidence of prior suicide attempts is also helpful. Toxicology results can be very helpful as well, particularly for suicidal overdose. And then injury characteristics, that I talked about.
There is no age limit for a homicide classification. If we see that a 3-year-old shoots their 6-year-old sibling, that is a homicide period. It is classified. It is a death as result of someone else. Most likely that child will not be charged with any sort of murder or anything. From our purposes, we would not think about as a homicide, that child pulled the trigger and shot somebody. We do not think about intent in the same manner we do for a suicide.
Exposure to violent content is occurring at younger and younger ages and kids are extremely impulsive. I talk about suicide as a permanent solution to a temporary problem. And that is a message I have tried to instill in my own children. Everything eventually, it might take time, will get a little better.
As we talked about and as Margy mentioned, 10 years old is the cutoff for the major data collection efforts and I think that is wrong. The best way to show there is not a problem is to not collect the data to prove it. If it is a suicide, you have to count it. It counts. An 8-year-old who kills himself counts. If we do not document it and we do not change that requirement, we are missing ways to intervene. If we do not know we have to go down into the elementary school and provide crisis counselors or mental health solutions if we are not going to make a change for these kids.
In the 10 to 24 age group as Margy showed, suicide is the second leading cause of death. This has increased. About 52.2 percent over the 2000-2021, according to the CDC website.
From a forensic pathologist standpoint, if you take the intent out of the picture, it makes it much easier. Did they shoot themselves? Yes. That is a suicide. But it is hard to not put the intent in there because that is by definition part of what a suicide is, an intentional act to kill yourself. But if you break it down to just did they kill themselves, that is a suicide. It can be very hard.
When my own sister killed herself, it was not 30 minutes before my family was saying she was murdered. I had to take my grieving sister hat off and put on my forensic pathologist hat and talk them down because someone out of the woodwork that I do not even know called my dad and said there is no way she would have killed herself. I read her suicide note. She killed herself. Just stop. It is making it worse. It does not help. These people that come out thinking they are helping the family, a lot of times really are not and they are making it worse.
Special circumstances are – I picked three special circumstances. Race and ethnicity is one of them. Proper data collection ensures that we get better statistics. We have to ask these questions. The demographic data usually is entered by the funeral home, which is entered by the family. That is not really something the medical certifier does. If that is incorrect, I can say that is their fault. But maybe they are not asking the question. It is wonderful that we are bringing in the funeral home directors and groups to address that potential problem.
And then we know that certain ethnic groups are at increased risk for suicidal behaviors. We have seen a little bit of a shift. It used to be that the suicide rates were typically higher in the non-Hispanic white individuals. However, that has flipped now. That has gone down but the non-Hispanic – the non-Hispanic black and Hispanic population demographic has increased in suicides. We have to capture this information so that we can address the prevention and hopefully decrease these deaths.
Veterans are another special population. Race and ethnicity, as we said, is captured on the death certificate. Veteran status is captured on the death certificate. But I am guessing a lot of times it is unknown because a lot of times we do not know. But this is information that can come from the family and the demographic part of the death certification. However, it is a part of our case management systems whether they are veterans or not. If you are requesting records from the VA Medical Center, they are a veteran. It is not rocket science sometimes.
Suicides are a leading cause of death in veterans under the age of 45. If you adjust it to the non-veteran population, it is about 57 percent higher. Again, excellent place where we can intervene and provide resources.
However, we have another group, the sexual orientation and gender identity group. This is not captured anywhere on a death certificate. That is not data that can be collected off of data certificate mining. This is very tedious data that is mined out of autopsy reports and death investigator reports, law enforcement reports, and case management systems. It takes hours. Unfortunately, what we find is that it is not collected properly. It is not done properly. The NVDRS group worked very hard to gather this information. But if we do not put it in our systems, they are not going to get it. Because it is not on the death certificate, we have to make adjustments and make it available somewhere in the autopsy report, in the investigative report, in our case management system.
It has been shown that transgender adults have a 40 percent lifetime prevalence of suicide attempts. And the highest risk is in the teens and 20s. It is a very complex topic though some family members do not want to talk about it. They do not want to admit that their child has a gender identity problem or sexual orientation that is different from what their traditional standards are. And perhaps that is the triggering event, the lack of acceptance.
There are a lot of unique challenges to this. If you have a transgender individual and we see physical characteristics of a male and they are identifying as a female, I do not know that. I only see the physical appearance of that person. That is a challenge for medical examiners and forensic pathologists to figure out how best to characterize that individual in our autopsy report.
I just came up with a few scenarios. It will be in my PowerPoint. But basically, they just show what we deal with every day. An 11-year-old boy watching YouTube videos with sexual content found with a soft ligature around the neck, partially suspended in a bathroom. No history of depression. No current stressors. This is probably not going to be a suicide. This is probably some sort of sexual experimentation. But again, you have to go and do the work to figure this out.
A 9-year-old obese girl found dead in bed with an empty bottle of sleep aid pills. She can buy those at Target or Walmart or wherever. There were notes indicating she was being picked on about her weight and not wanting to live anymore doodled and scribbled in her journals and her classwork. I would classify that as a suicide. She is clearly having these thoughts and feelings even though she is young. We cannot take – just because they are less than this age that they cannot have thought about killing themselves.
This one was really tragic one. This is similar to one case that I had. He is a 17-year-old boy, well liked, going to the Naval Academy, good relationship – going to college, found hanging from a tree. But the day before and this was wonderful that this boy felt comfortable enough to tell his mom, “Hey, I sent an inappropriate photo to a girl and now I am being blackmailed by this girl and they are going to send it out to everyone including his college if he does not pay her money.” She was like do not worry about it. We will take care of it. It is going to be okay. Do not pay the money.
He paid it anyway. They demanded more money. He did not want to bring shame on anyone else and he killed himself. These triggering events are so helpful to understand what was going on in that person’s mind but often we cannot see that on the basis on the autopsy alone. The death investigation piece is critical to us. Thank you.
JAN GORNIAK: Panelists, please turn on your cameras and then we are using the ceiling microphones so you can just talk as you --
LISA COLPE: Jan, if you have some things that you would want to start out with in terms of your summary – statement things and then we will go into questions and answers.
JAN GORNIAK: Good morning. Good afternoon wherever we are. I do not even know anymore. Jan Gorniak. I am a forensic pathologist and medical examiner at Clark County, Las Vegas. One thing I wanted to say is repetition is the key to mastery. That is what I learned in medical school. Repetition is the key to mastery. We heard things being repeated, the cause of death, manner of death, the things that need to be collected. So I am honored to be here.
When people ask me what do I like about forensics, what type of case, I love forensics, but early in my training, I found I was drawn to suicide in vulnerable populations, the pediatric populations, the maternal, and also the incarcerated population. I am so happy to be sitting here. It sounds kind of weird but thank you.
The initial case I had when I was in fellowship training was a 17-year-old black male who was found hanging from a tree. First of all, it was not a suicide. As an African American female, one of the myths are that black people – we do not kill ourselves especially not hang ourselves because of the correlation to lynching. You will find that a lot – people will start saying it is not a suicide especially if it is a black person hanging. This young man was found in his yard. This was in 2004. He had an online journal.
At that time, young in the profession, I used to read through all of the notes, not anymore. I cannot do that. I do not like videos either unless I have to. But I sat there and I read his journal. I can tell he had some depression going on just reading it.
One of the things that really hit me was he had done an English paper. He wrote an English paper and it was titled “why life sucks for me”. What did the teacher do? The teacher made him rewrite it. There was no intervention at that point so he wrote a paper on frivolous lawsuits. That really struck me because especially with children in contact with teachers and that was an opportunity for an intervention at that point and it was missed.
We talk about this in our other groups that I raise my kids differently now or did. They are grown. I do not have to raise them anymore. Because of what I was seeing. A young man. I think he was 14 that got caught shoplifting. Mom had to go and pick him up. Probably had the speech in the car. Go up to your room. Wait for your dad to get home and then dad gets home, he is found hanging in his closet.
When I used to send my kids to their room, it would be for five minutes if that. No. No. No. You come and you sit right next to me because you do not know what those triggers are. That is different. I have a son and daughter both raised in the same house but respond to things differently. I know what a trigger would be for my son compared to my daughter and I think that is very important.
COVID. We have mentioned. Not laying eyes on children in school as a big factor because you could say Joyce does not look too good today. Is there something going on? But we also know during COVID you heard that domestic violence reports increased. But the child abuse reports were decreased. Why? Because they were home with their abusers.
I had a young lady on Zoom and she had filed a sexual assault complaint against one of her classmates. Who did she get to see every day on Zoom? Even though they were not in the classroom, she was seeing him every day on Zoom. They reported it. It was not substantiated and then unfortunately, she ended up as one of my patients. You talk about means. The hangings are usually the most common means that young people can get to but also Amazon Prime. Sodium nitrite – I am starting to see that a lot. You are just like – you can buy it. Make sure you have your parental locks on your Amazon Prime so they just do not buy it. It will come in the mail. You never know.
Talking to families. That is one of the most difficult things to do. Some families are “accepting” of it or understanding, and some will not believe it. Some of the things I – when I speak to families, I have to say I do not know your son. I did not know your husband. The intent. This is a medical, sterile diagnosis I am making. But you know them. They can believe whatever they want to believe. This is not a judgment on their child or anything like that.
When we were talking about the intent or the letters of intention, not a suicide note, trying to look everywhere, those Facebook pages, those text messages. They are around. You have to get all up in people’s business. It is not easy. I think this is the second most difficult investigation that the medicolegal death investigator has to do behind the child, the pediatrics, the infant death. But you have to ask those questions and it is very important like Lauri says. If you do not ask, you are never going to know. The sooner, the better.
I had a conversation with my husband the other day about use of pronouns. Do you ask people the use of their pronouns? I think that is another thing that Lauri was mentioning. You cannot be afraid, and you cannot assume. Ask. And then unfortunately, we also have to believe what we are being told. Everybody is not forthcoming.
I had a case when I was in Atlanta where we found out more information based on medical records that the family was not willing to share, not even with the police. That made our determination of suicide. Even though the family knew, they did not want to share.
I think what we are talking about right now and I cannot wait for the rest of the day and tomorrow is I say so what. We investigate these deaths. We have the data. So what? That is where we need to get to. Obviously, prevention is first and foremost. Now that we have the data, now what is our next step so we do not have 8-year-olds dying by suicide?
When I trained, like I said in 2004 and 2005, we would not rule a death of a child under the age of 12 a suicide whether it was a gunshot wound, whether it was a hanging. If they were less than 12 years old, we would not do it. I believe there are some jurisdictions out there still that have an age cutoff.
I recently had a discussion with a pathologist where they said – the young man was 16 years old. His sister and his sister’s daughter were in the room also. He shot himself. They wanted to rule it an accident because he was not unfamiliar with guns. But that does not make it not a suicide.
My response back was what if he had shot the little 3-year-old? Would you have called that an accident because they were not familiar with guns? Like Dr. Pinneri was saying, it is the act, not the intention behind that act. We would have no problem calling it a homicide if he had shot a 3-year-old. We shouldn’t have an issue calling it a suicide because he is unfamiliar with the weapon. We have to be very objective when looking at these cases.
LISA COLPE: Does anybody else have any questions?
JAN GORNIAK: I do have some here. Is religion or sect captured in the demographic statistics in the United States and on the death certificate?
PARTICIPANT: Currently not.
JOYCE DEJONG: Do any of our offices collect that?
PARTICIPANT: We collect that for suicides and drug overdoses as part of our interviews with family.
PARTICIPANT: And Michael, who are we? You said we collect it.
PARTICIPANT: Office of the medical examiner.
PARTICIPANT: I did want to pick up a little bit on what you were talking with the age cutoff because that has come up a number of times in our discussions during our planning of the committee. It has been raised to me because of this distinction of 5 or younger or 10 or older and you are saying 12. I have heard you said 16 as well.
I know that during our other panels, some of the researchers have – during the other NIMH panels we had, the researchers were talking about their better understanding of children’s ability to understand the consequences of their actions. I think I heard if I am not mistaken, Jan or Dr. Gorniak, – and this idea of intent – I laid out a bunch of different things to talk about. But the main one was the idea that there is some sort of age cutoff, which I think is pretty common – there are certain medical examiner’s offices that have – it is either by tradition practice or whatever, but also within the public’s eye because we have people in the public with some consideration.
JAN GORNIAK: Those are the ones too – I had an 8-year-old and that was actually – the media. The news wanted to put that on as a prevention but I get it. When children use the words, I am going to go kill myself, why don’t we believe them regardless of how old they are? We brush that off. I have three young grandbabies who – their tablets die. That is the words they use. My tablet died. I am going to plug it back in. They get mad. I hope you die. Do they really understand? I do not know but those are the words that they are using and what they equate it to. Eight is the youngest I have ruled a suicide. I am not sure about anybody else but that is the youngest.
JOYCE DEJONG: We had an 8 – I do not know that I have not had one before. I think too because in that age group, it is not – you saw the numbers that Dr. Warner presented in the beginning. There are not high numbers and these are situations that need to be taken on a case-by-case basis without some blanket – they are 8 and our cutoff is 10 so it must be something else. I do not think you can do that –
It is also in that age group where you see a lot of death investigation systems that lack objectivity and in the interest of not hurting the family’s feelings, will not put suicide on there. I advocate for death certificates that just are proof of death and if they want to have the whole thing with the document, that would be good. That way the family does not have to carry around something that – to the bank and to turn off the consumer’s power or whatever that says he did this. There is too much data that gets distributed. I just need to show you that my family member died. You do not need to know why and what happened but yet to actually advocate for individual cases and we cannot – I do not like to hurt people’s feelings. I wish I could say it was not whatever it was. But you have to be objective on that document.
KATHRYN PINNERI: Along those lines, I think that we found that depression and mental health disorders can be genetic and they run in families. You are actually doing the family a disservice by hiding that bit of information by not providing, hey, your Aunt Susie killed herself. You are taking away an opportunity for prevention and help and intervention for the family members that come behind because that is a document that is a forever document. It is going to be on Ancestry open source in 20 years or however many years it is. To not give that information I think is doing a disservice. I do not want necessarily to have that on the death certificate for people. It is very sensitive. But it is also something to talk about and not hide and share and prevent.
JAN GORNIAK: Not only do mental health runs in families, suicidality runs in families too. And then you have your survivors of suicide. Then those family members are now at a higher risk of dying by suicide because of the violent death that they “survived”.
I have some questions here but I do not think that they are --
PARTICIPANT: I have a question that you could answer by –
PARTICIPANT: We wanted to talk about age and understand that spectrum. But the difficulties you encounter with determinations I’m thinking about social media. I am wondering about how challenging that is. Are you able to think about it in all cases or in 5 percent of your cases – things that you are struggling with just in suicide determinations in general. What are the biggest things that you are running into?
KATHRYN PINNERI: I think the electronic and social media almost are my number one go to now because that is how it comes. Somebody shows up at the house – I had a text message that says blah, blah. I love you. It has been nice knowing you. Good-bye. It is not a suicide note but that certainly is suspicious sounding. That is where we are getting all this information from.
JOYCE DEJONG: It is challenging to get, especially if you do not have access to their social media websites.
KATHRYN PINNERI: It has to be forthcoming, and it has to be something that you can access. That is not always the case. There is another end to that message that went out. Sometimes we get reports from somebody in Kansas who received a message on Discord about something and they show up at someone’s house and say this was reported. That is wonderful. That is a great tool that kids can use to help other kids out. The family who are just sitting there having dinner and the police show up. Oh my. What is going on? But access to that is very tricky. Most parents do not know what all their kids are on. Some of it is designed so that they cannot see what is out there. It is very hard. But it is my number one thing for teenagers.
JAN GORNIAK: Same with ours. You have to get access to it. And some family members are very forthcoming. I had a case when we talked about undetermined. It was a transgender female. I think she was 16. But everybody wanted to say the family did not – the family was very accepting and very engaging with her. But was it an auto erotic versus a suicide? We could not get any of that information. We sat on that case – I think someone asked how long does it take. We sat on that case for a long time because there are some people who do psychological autopsies. We tried to do that, but there was not enough information.
I finally ended up reaching out to the mom and then the mom had found photos, prior photos of auto-erotic behavior. That was able to help. But I was leaning toward suicide for – but without that, that would be an undetermined death. Was it a suicide or an accident? You have to get access to that. And sometimes, as investigators, we have to ask. We have to ask for the phone. We have to ask those difficult questions.
JOYCE DEJONG: There is a question on here that I am going to quickly address and that is how come there is not as much attention paid to the family members and does that increase the chance of suicide. I alluded to that earlier. It is resources. I would so love to every time there is any death but especially a suicide to have professionals that could respond but we do not have – as the medical examiner, we do not have the resources for that and we do not have in the counties to say let us send our suicide prevention team out there to deal with this family. It is resources.
KATHRYN PINNERI: I worked in an office with one and it makes – with a social worker, and it makes an amazing amount of difference. And even just a death notification – if it is done with a chaplain of their religious preference, that is so helpful for the family. You are starting out at a different point, a different place where they feel like they have some support and that goes a long way. But without the social worker, we really just do not have the bandwidth ourselves to handle that.
LAURI MCGIVERN: A question on here for me specifically about why public health nurses specifically for child deaths versus social workers. Again, this is just in my jurisdiction. We use public health nurses from the maternal child health groups specifically in infant deaths. These are often unsafe sleep. It is an opportunity for us to provide some education around that and the public health nurse can also talk about resources if they need a crib or for future children or other children in the family. We use the public health nurses in that sense, but we have used the social workers from our law enforcement for other violent deaths, which has been extremely helpful.
LISA COLPE: I know we are getting ready to wrap up. Thank you so much. Can we ask questions?
STEPHEN O’CONNOR: (Inaudible)
PARTICIPANT: When you answer the question, can you please recap that? We could not hear Steven online.
KATHRYN PINNERI: Just a quick recap. Are there ways to improve the racial determination or ethnicity determination for these high-risk groups to get --
STEPHEN O’CONNOR: Determining the manner of death - disparities in how accurately it is done for different race and ethnic groups. There are factors that perpetuate that disparity. That is a major concern, especially when we see rising emergent of suicide among certain ethnic groups. Alaskan Native, it looked like that data said 34 percent (inaudible). Can you all speak? What factors do you think lead to that kind of disparity among certain groups?
KATHRYN PINNERI: I think some of that disparity – we do not enter that data. That is not our data for the death certificate. That comes from the funeral home and their conversations with the family. If they are not collecting that properly, that is the problem for those groups.
JOYCE DEJONG: It is a really complicated question actually that we probably cannot – that is a multi-factorial issue that –
KATHRYN PINNERI: We were talking earlier that – Jan and I were talking about why that is because that is information that comes from the family. But the only thing we could think of is maybe the funeral homes are not asking those in-depth ethnicity questions. They are not getting down to that list. Perhaps it is a form that the family should fill out before they get there or maybe not do it verbally so they are not running through the 12 different things while the person is grieving, and their attention is not what it should be.
But I do know that there are groups and individuals that have offices located in these jurisdictions with high populations of Native Americans or Alaska Natives. They would be more familiar with the how and whys those groups have increased.
STEPHEN O’CONNOR: I know that is complicated. Maybe we could pick it up later, but it is a major concern. There are rising rates of certain groups but there is a lot of indecision and determined fear of death. That is something that --
KATHRYN PINNERI: I also think that that is a data collection thing as well, because now we know how important this data is to collect and we are collecting it. So maybe it is that it is being collected now and that is why we are seeing it because they are actually asking the questions. There are more boxes on a death certificate now for these items. Maybe that could also be – that the information is just being collected now whereas 10 years ago, it was not being collected. They were all just classified as one thing.
MARGARET WARNER: I did that statistic, and it is older and after we have learned about the – there have been a number of efforts to improve the quality. And some of it was about self-identification versus what is on administrative records. It is very complex, but ultimately the people cannot ask them to self-identify their race.
LISA COLPE: Thank you very much to all of you.
BRENDAN WEINTRAUB: Good afternoon, everybody. Welcome back. Just like we did during our last break, we are going to come back with a quick Mentimeter. I am going to go ahead and share that right now and share that to the room as well. Similarly, you will go to www.menti.com. That same code, 31733. And our very ambitious question that we wanted to ask folks is what research-to-practice or practice-to-research questions should be prioritized to address youth suicide. Just take a couple of minutes here. As folks put in their answers, they will be popping up on the screen – with some of these answers so far. Some rigorous evaluation of prevention programs, how best to include schools and data collection, the impact of social media, wraparound services. These are all great. About social media again.
You can continue to submit as we are moving on to our next section now. We just wanted to get some folks engaged into what you are thinking about in terms of these types of questions of course. We have our presenters and panelists today that are going to discuss some of these as well. I am going to go ahead and stop sharing my screen.
I will pass it over to Dr. Jane Pearson, who is going to begin our next discussion.
Agenda Item: Overview of Research on Child and Youth Suicide; Epidemiology and Prevention Considerations
JANE PEARSON: It is a real pleasure to help be the discussant for this panel of folks. I do not think I will have to do too much to get the conversation going. A number of these folks – I think all of them have presented at NIMH conferences on child and youth suicide already. You will find out reasons why we have asked them to come back and talk some more because I think they will be covering topics that are directly relevant to some of the issues already raised.
I do not want to take a lot of time. You have seen the list of speakers here. Their biosketches are on the web. I am just going to turn it over to Dr. Joan Luby right now. Go ahead, Joan.
JOAN LUBY: Thanks, Jane, and thanks for inviting me. Just to rush right into it, following some of the discussions we were having this morning about the 8 and under cutoff, I think we can clearly establish that a cutoff at age 8 or 10 is definitely the wrong thing to do and we are going to miss a lot of our signal if we use that cutoff. It seems like there is a lot of agreement on that.
Along that line, what I am going to talk about are signs of suicidality in children between the ages of 4 and 7 and to talk about how we stumbled upon that in our studies of preschool depression. I sort of became an accidental suicide researcher because I have been studying for much of my career what clinical depression can look like in younger children particularly in preschool kids and establishing as have other labs that clinical depression can arise as early as age 3.
It looks a lot like the adult disorder, similar DSM characteristics, but it has developmentally translated symptoms. Kids with preschool onset depression have some of the same biological markers. They have some of the same neural alterations. And we followed a sample of about 300 preschoolers now for almost 20 years and have shown that they go on to have recurrent rates of depression. They have alterations in brain development, et cetera. We know this is an important clinical phenomenon.
When we first captured this sample in 2002, we were not particularly interested in suicidality. There might have been a couple of questions about suicidality. But we had a very, very low prevalence rate in our population of preschoolers. We do have some publications looking at how it evolves across development, but then when we captured another sample more recently in about 2016, a clinical sample, we were finding these very startling rates of suicidal expression and we decided we needed to really drill into this and we got an NIMH supplement to ask more questions about preschool suicidality.
What a depressed preschooler looks like and why they are often missed in clinical settings of course is because they are internalizing. They do not cause disruptions in a classroom. But they have symptoms like anhedonia. They feel like nothing is fun. They feel excessive guilt. They feel guilty for transgressions that they committed or transgressions that they did not commit. They feel like kids do not like them. They feel like they are bad and they have passive suicidal thoughts like I wish I was never born.
But in this new sample of preschoolers, we found that they were also expressing that they wanted to kill themselves. Of course, when a 5-year-old in a school expresses they want to kill themselves, there are two reactions. Either adults just want to ignore it and write it off or they get very, very panicky and they send them to the emergency room. It is very – clinicians and mental health professionals and medical professionals need guidance on how to address this problem.
In this new sample of depressed preschoolers – this was a sample of about 230 depressed preschoolers who were engaged in a randomized clinical trial of psychotherapy for depression. We found these really high rates of suicidal expressions. A lot of kids were expressing passive suicidal ideation, things like I wish I was dead. Kids were saying I want to kill myself and then some kids were displaying suicidal behavior: wrapping things around their neck, jumping from a second-story court, running out into traffic. This was very alarming, not something that we expected. We attempted to ask more questions about this.
I think it is fair to conceptualize suicidality really across development as a feeling of intense internal distress that is intolerable and that intolerability is what causes people to come to these rash behaviors to solve that problem.
Going back to the question of whether young children could express suicidality and the idea that this was not really valid. These kids did not know what they were saying. They really did not understand what it meant to die. In fact, and this is sort of a trend in developmental science, we are wrong when we make those assumptions about young children. In fact, 4- to 6-year-old children who express suicidality had a more mature understanding of death relative to their healthy peers and to their depressed peers without suicidality. I will show you in a minute just how we assess that because we had some developmentally appropriate, fairly detailed death interviews.
The other thing that we found is the children who expressed suicidality depicted more violence and more suicide in their conflict resolutions when we gave them story stems. Story stems are something we use with young children to give them probes. There is a conflictual story and they have to complete the story. These kids who had suicidality were completing the story using violent themes or people dying or people killing themselves. Both of these findings are published. It is the work of one of my mentees, Laura Hennefield, who has since gone on to get a K Award to really study the issue of suicidality in young children, what it means, how it is expressed, what the antecedents are, et cetera. I am very much looking forward to some of the things she comes up with.
But when we get to the question of whether suicidal young children understand death, we see from this, looking at detailed interviews about their understanding of death that children who have suicidality understand death better. They see that it is more – they understand the universality of it. They understand the irreversibility of it. They understand the things that cause death, et cetera. It is definitely a misperception to think they are saying things and they do not understand.
Although I very much appreciate the things that were said – sample of children who had these high rates of suicidality in the preschool period and assess them in early adolescence.
One of the things we have found is that having suicidality in the preschool period is a very robust predictor of having suicidality in the pre-adolescent period. You have a seven times greater odds of having any kind of suicidal thoughts and behavior in pre-adolescence if you were depressed as a preschooler. You have a six times greater odds of having active suicidal ideation if you were depressed as a preschooler and seven times greater odds of having suicide attempts if you were depressed as a preschooler.
This goes for occurrences since preschool or within the last month of when we assessed you when you are pre-adolescent. This would go to show that having preschool onset depressive symptoms or the clinical disorder is a robust risk factor for suicidality in early adolescence.
I already mentioned that this was a follow up of 137 of the kids who participated in this randomized controlled trial. We followed them at ages 8 to 12. And we also had a peer control group or community control group. This just summarizes what I just showed you, these very robust odds ratios of having STBs in early adolescence, which is the time when the general public accepts that this is a very important risk factor.
Importantly, these risk ratios arise even when we control for externalizing comorbidity, also a risk factor for suicidal thoughts and behaviors such as ADHD and oppositional defiant disorder. It also arises when we control for family history of suicidality. We know suicidality runs in families. And similar outcomes arise in depressed preschoolers who did not have suicidal ideation. You are at high risk if you were depressed as a preschooler and had some suicidal thoughts or behaviors. But you are also at high risk if you did not have those suicidal thoughts or behaviors. This is just a graphic that sort of shows you the differences that give rise to these odds ratios between depressed preschoolers and their community controls with much higher rates at pre-adolescents in depressed preschoolers.
Again, looking at the continuity, here, we took depressed preschoolers, and we broke it into two groups, those who had suicidal thoughts and behaviors in the preschool period and those who did not. You can see both groups have higher risk of later suicidal thoughts and behaviors. Even the kids with depression without early suicidal thoughts and behaviors go on to be a much higher risk for these things in early adolescence as do the kids who had suicidal thoughts and behaviors in early adolescence.
The other thing we wanted to take a look at is how parents and children agree or disagree when they report on suicidality in young children. Because of course, when we are assessing 4 and 5-year-old children, we cannot sit them down and conduct a psychiatric interview because that does not work very well. We tend to use other methods. We use things like these story stems. Some people use puppet interviews. There are some things that children can and cannot reliably report on and sometimes we rely more on the parent report or caregiver report.
What we know in mood disorder research in childhood is that parents and children very much tend to disagree when they report on these symptoms. We basically find the same thing when we look at suicidality in early childhood. Parent and child reports do not match very well, which again really suggests that it is important to assess these symptoms directly from the child. The parent does not necessarily know about all of these symptoms.
Now, what I want to do is to emphasize what might be one modality as a preventive intervention. The clinical trial that we tested in the sample of depressed preschoolers was an intervention that is called parent-child interaction therapy emotion development. It targets depression as a disorder of emotion development and it utilizes the parent as the child’s external emotion coach and emotion regulator.
The therapy is done with a therapist on the other side of a one-way mirror, the parent and the child interacting together, and the therapist is coaching the parent during a live in vivo interaction, often with mild stressors. This treatment proved to be very effective for the treatment of depression and had robust effect sizes of about 1.0.
What the emotion development module does is it targets the child’s ability to accurately label their own emotions to identify the emotions in others to learn to regulate intense emotions but to experience an intense emotion, even an intense negative emotion. We view it as healthy to be able to experience and tolerate that emotion but to be able to regulate it.
It did not target suicidality per se but it does target the ability to regulate and tolerate the stress. In that sense, I think it is an interesting preventive target.
What we use in the treatment if we use evocative events in vivo during live coaching. The parent importantly, is taught to tolerate the child’s expression of intense emotion. We teach parents to stop doing things like saying don’t cry. That is nothing to cry about. The parent is then taught to recognize the child’s emotion, validate the child’s emotion, and help the child regulate the emotion.
And the reason why I am telling you about all of this is because the treatment and particularly this ED module very much changed the way as planned as hoped for – changed the way the parents parented their young children. Parents tended not to minimize their child’s emotions when they learned this ED module. They tend to encourage expressive expressions of emotions after they learned this ED module. And they tend – distress reactions. In other words, parents are learning to validate children’s emotions, to encourage more emotional expression, and to help children regulate these emotions.
The reason why I am telling you all of this is because when we looked at how treatment impacted the later prevalence of suicidality in pre-adolescence, we see that the children who achieve remission as a result of this treatment during the preschool period, which is a 12-week treatment delivered by a master’s level clinician, those kids are significantly less likely to have suicidality in pre-adolescence, which I think is very interesting and promising because the treatment did not directly target suicidality and the treatment could certainly more directly target suicidality. But what it does target is emotion regulation, emotion tolerance, and essentially emotion intelligence. In that sense, I do think that a preventive intervention starting earlier in childhood that targets emotionality and emotion regulation could be a very promising intervention.
Just to conclude, what I am trying to argue here is that what might be very important is to pay more attention to early onset mood and depressive symptoms because I think this is a robust marker for later suicidality. This holds and has powerful odds ratios even when other known predictors are controlled in the analysis. Depressed preschoolers without suicidality still had a higher risk of suicidality in early adolescence. And this early intervention that modifies how parents manage the child’s emotion in early childhood decreases the risk of suicidality in early adolescence. The results suggest the clinician should attend to early STBs and the early onset depression might be an important marker.
I guess the other thing that I did not add to this slide that I have been thinking of course, the pregnant question that everyone is asking is why are these rates of suicidality increasing over time? Why are they arising younger and younger? And I know that there is a lot of very interesting program announcements out from the NIMH to try to investigate this issue. But just standing back 10,000 feet, I think we are living in a society where we have increased rates of intolerance, increased rates of bullying, portions of our society where bullying is valued and revered. And we have increased exposure to violence and increased exposure to lethal means. That grew certainly to me seems like a very important etiology.
With that, I will turn it over to Arielle, who is the next speaker.
ARIELLE SHEFTALL: Hello, everyone. I am going to take a little bit of a step back and talk about the trends that we have been seeing with youth suicide and suicidal behavior. And before doing so, I do receive funding from the National Institute of Mental Health as well as the American Foundation for Suicide Prevention. I do not have any other conflicts to disclose.
As stated, I will just review the epidemiology as well as talk about some risk factors.
Before getting into the nitty-gritty so to speak, I really want to put this slide up. I think it is the most important slide out of my entire deck. Every presentation I give I always have it up and the reason being is that I believe that every single person in this virtual room as well as in the room can actually do something to prevent suicide and suicidal behavior. I always encourage people to have these numbers in their phones because you never know when yourself is going to need them but also someone else that you know is going to need them. If you do not mind taking a picture real quick and again the hotline number is now three digits, 988. We can call that number. We can text that number. And then the crisis number is a text line that has been available for quite some time now. Again, I think everybody can actually play a role in preventing suicide and suicidal behavior and these numbers can help do so.
Also, I stole this slide, I have to admit, from Dr. John Ackerman at Nationwide Children’s Hospital. But I think it really speaks to the fact that we have to care for ourselves. This topic is very, very, very heavy. And no matter which piece of the puzzle you play whether you are a clinician, whether you are a medical examiner, a coroner, a researcher like myself, talking about suicide, looking at cases all the time is very, very hard to do. It is hard work to do. It is necessary but absolutely hard. But I do want to encourage folks if during this conversation that we are having, my specific slides, if you need to step away, mute me, get out of the room, do what you see fit to take care of yourself because that is truly important. This work is hard. It is needed. But sometimes taking a break is necessary so that we can come back even stronger and do this good work and to do it well. I will not be offended in any way, shape, or form if that is what you need to do.
These are just some definitions. I am going to be saying these words a lot. They will be said throughout many of the presentations and some of you are very familiar with these definitions but just in case some are not. Suicide is under the umbrella of suicidal behaviors. Suicide is the death of an individual. It is caused by an injury to oneself and there is explicit or inferred intent to die that is present.
Suicidal ideation is thoughts about suicide and those can range. Those can be very passive where you just wish you were dead, wish you would go to sleep and not wake up, all the way up to actually having a plan with intent.
And then suicide attempt also falls under that umbrella of suicidal behaviors and that is the non-fatal self-injurious behavior to oneself with that stated or inferred intent to die.
Another term that is stated within the field is non-suicidal self-injury or NSSI. This is not under the umbrella of suicidal behavior because the intent is not present. With non-suicidal self-injury or self-harm behaviors, another term that you will hear, this is when someone actually hurts themselves on purpose physically but again without that intent to die. Usually people that do engage in non-suicidal self-injury are doing so to either get emotions out or to actually feel something. Those are some of the reasons that you will see in the literature.
With the problem of youth suicide, in 2021, suicide was the tenth leading cause of death for all age groups. However, it was the third leading cause of death for youth 10 to 19 years of age. In that timeframe, it accounted for over 2900 deaths in that age range. This was actually more deaths for eight leading causes of death in this group.
When you look at suicide by sex, what you will see is that both males and females have seen a significant increase over time. Overall, we have actually seen a 45 percent increase in the past 21 years for 10- to 19-year-olds in the US. For males, we have actually seen an increase of 28 percent but our females have actually seen the largest increase. From 2000 to 2021, we actually have seen a 129 percent increase in the rate of suicide for girls 10 to 19 years of age. This is about if you are doing the math like 4.6 times higher than the increase that we have seen in males.
When you look at self-harm behavior, we have also seen an increase. And the way that self-harm – the rates of self-harm are actually looked upon in the nation is through the CDC. And what they do is they look at ED presentations for self-harm behavior. And the data goes back to 2001. Right now, I think the data is up to 2020. And then for 10- to 19-year-olds over this timeframe, we have actually seen an increase of 181 percent in the number of presentations of ED visits for self-harm behavior so the youngest of the young.
This is an area that I focused most of my research upon. I do research looking at youth 6- to 12-years of age and looking at early vulnerability. And the reason being is that the leading causes of death has actually changed drastically in this age range.
For our 5- to 12-year-olds in 2009, suicide was actually the eighth leading cause of death, accounting for 82 deaths in that specific year. In 2014, it rose to the sixth leading cause of death. And then in 2019, it actually rose to the fifth leading cause of death, accounting for double the numbers of lives lost than it did ten years prior.
In 2020, it remained the fifth leading cause of death. It accounted for 180 deaths. And the most recent data that we have right now, which comes from the CDC WONDER database, is for 2021, where it continued to be the fifth leading cause of death and accounted for six more deaths than what it did in 2020.
And then when you look at the suicide rate over time. I started this slide at 1990 as you can see right here, it goes up to 2021. The line actually is the rate of suicide whereas the number of deaths are actually these gold bars. And what you can see is from 1990 to 2021, we have actually seen an increase of 204 percent in the number of suicide deaths for 5- to 12-year-olds.
Good news if there is any, from 2018 to 2019, we actually did see a decrease. That is real. We saw about 11.4 percent decrease amongst these 2 years. However, from 2019 to 2020, we actually saw a 10 percent increase. And then from 2020 to 2021, we saw a 4 percent increase, so going in the wrong direction.
And then when you look at self-harm behavior among youth 5 to 12 years of age, again, over this 19 period of time, we have actually seen an increase of 411 percent in the number of presentations in the emergency departments for self-harm behavior in youth 5 to 12 years of age.
When you look at mechanism – this is looking at all of the youth so 5 to 19 years of age. What you will also see is that we have seen changes over time. For females specifically, the leading cause of suicide is hanging suffocation. However, we have actually seen large increases in suicides by firearms as well. Over this period of time, just the 10 years, we have actually seen an 80 percent increase in the number of firearm suicides by females. As many of you know, hanging suffocation and firearms are considered very lethal means of suicide and they account for 79 percent of the suicides seen in this ten-year period of time for 5- to 19-year-olds.
For males, we have seen increases in firearm suicides as well. It was 59 percent over this 10-year period of time. And the majority of males who died by suicide do by firearm, then second to hanging suffocation.
And then when you look at the data in 2021, we actually see again another increase present for our females. For the percentages – 22 percent for firearms over this 10-year period of time and just in 2021, it was 24 percent for firearms for females, 53 percent for hanging suffocation. And then the firearms for males actually was at 57 percent followed by hanging suffocation by 31 percent.
Risk factors. Many of you that are on the call know what a risk factor is but just in case. A risk factor is just something that increases your chance of a behavior happening. In this specific instance, we are talking about suicidal behavior. And some of these risk factors are changeable just like Dr. Joan Luby just talked about emotion regulation. We can train parents to help kids regulate their emotions better and give them those skills whereas others are not. We know that family history of suicidal behavior increases your likelihood of also engaging in suicidal behavior. That may be a genetic component present for suicidal behavior.
And others also depend upon your personal circumstances so your age, your sex, your gender identity, your sexual orientation, your race, your ethnicity. Risk factors are dependent upon your environment as well, where you live. When we get most concerned is when we have multiple risk factors occurring at one time.
And some of the known risk factors for suicide being male, access to lethal means, LGBTQ+ sexual orientation and/or gender identity, having a history of sexual or physical abuse, and many, many, many more.
We have also seen in the literature that social media has been associated with suicidal behavior but also I think there are ways of engaging with social media in a positive manner to use social media in terms of interventions. I think unfortunately it is a double-edged sword. It does increase the risk, but also I believe it can actually decrease risks if used appropriately.
When you look at younger kids, there are different risk factors that have been shown in the literature, impulsivity being one, cognitive skills, as well as emotion regulation. Impulsivity, just having the lack of coping with whatever difficult situation they are experiencing at that specific time.
Research that I have conducted with colleagues when we compared early childhood suicides to early adolescent suicides, we found that kids who died by suicide were more likely to actually have an argument with a family member or a friend and they just went and acted on that upset-ness or those emotions that they were experiencing versus kind of processing that information and using the emotion regulation skills that they may or may not have.
Also, cognitive skills. Less likely to imagine the future. When I speak to my 6-year-old, she can only think about the here and now and that is absolutely where she is supposed to be. She is 6. If you look at the development of cognitive development, that is normal. She, when she experiences a really painful experience, may not see that that is going to end. She cannot see that there is something past the experience that she is currently having. Unfortunately, we know that that can be a risk factor for younger kids when it comes to suicide and suicide behavior. And then again emotion regulation, which Dr. Joan Luby did elicit in her presentation. Not being able to have those skills to regulate one’s emotions can actually unfortunately lead to these behaviors.
Last but not least, I just want to give a shout out to all of my family and friends that have helped me to do the research that I have done. Now, I am going to turn it over to my colleague and friend, Dr. Rhonda Boyd, who is going to talk to us about something else. Thank you so much.
RHONDA BOYD: I will be talking today about the challenges in ascertaining suicide risk in child and youth suicide.
For disclosures, I have no financial conflicts of interest. I do have funding from the NIMH and these are the funding sources that they come from.
Today I will be covering a range of topics. Many of those have already been discussed previously but that’s what I will be bringing together some of the topics and highlighting them further. I will be talking about youth/caregiver discrepancies in reporting, stigma, limited information available for younger children, and racial misclassification.
Some of the issues when you consider assessing suicidal ideation and behaviors. This came from a special issue discussing the discrepancy literature, focusing specifically on youth suicide. And some of the things that came up is what exactly is asked when you are assessing these behaviors, the content. What are people saying – the types of behaviors as we say there is a difference with suicidal attempt, a suicide plan, a suicide behavior? Exactly what you are asking may impact the response that you receive.
And how assessments are administered, the format. They can be asked verbally. They can be asked by questionnaires. These types of formats impact the response that you may receive. And then when measures are administered, their time interval.
Sometimes when we ask about suicidal ideation and behaviors, we are asking about maybe last week. We could be asking about someone’s lifetime. Your ability to remember that time interval impacts what response you may receive.
And then who is asked, the informant? It has been brought up about asking parents and what parents may know. In this case with a suicide death, we are not supposed to ask the child. That piece is maybe missing and there maybe other information, ways to get that as has been talked about maybe through social media or journaling or things like that. And then also, we have another circle of people that come in contact with youth - teachers, peers, and other family members besides parents.
In the discrepancy review literature and it has been brought up that it is very frequently that particularly parent and youth do not agree. Oftentimes parents and teachers and youth and teachers even see things differently. But when we think about suicidal ideation behavior, they similarly may produce conflicting information. Separate questions could yield different reports from the same person. That can impact how a question is raised.
Also, collecting information from multiple informants could yield very different responses. A parent may tell you something different from a peer about a youth as well as a teacher. We need to pull all this information together. I think the first panel when they talked about all the different information they pulled together, it is an example. What does all this information mean? How do you make a coherent story with these different reports? This is what the discrepancy literature reflects on.
I am going to present a study that my colleagues and I did in Philadelphia, looking at a large database of youth between 11 and 17 who reported on suicidal thoughts of killing themselves as well as thoughts of death. We looked at this discrepancy. I know Joan Luby talked about research with younger kids. These are slightly older youth.
What we found is that actually that disagreement is pretty common. In particular, there was a group of what we call unawareness where the adolescents said yes. They have these thoughts and the parents said no. For thoughts of killing themselves, there was 50 percent of the parents did not know that their child in their lifetime have thoughts of killing themselves and wishes of death was even higher, 75 percent. In this case, parents were mostly not aware that their teenagers were having these thoughts.
We also looked at what are some correlates to this unawareness for parents. Thoughts of killing ourselves, having a younger age. Younger age, parents were more likely to not be aware and this is sort of when we were talking about these younger youth, parents likely may know.,
Also interesting is that older boys particularly, they did not know. Fathers were more likely to not be aware of these thoughts as compared to mothers. And also having a history of therapy and psychiatric hospitalization actually made parents more aware. The were more likely to be aware that their child was having these symptoms.
When we looked at thoughts of death, actually for racial minoritized populations, parents were more likely to be unaware. However, for Latino ethnicity, parents were aware. It varied based on race and ethnicity. And then a family history of suicide, which has been brought up previously and in some of the chats. When there is family history, parents were more likely to be aware of thoughts of death. I think that point did come up. Knowing about your family history makes a difference because these families are likely to look for this with their children.
Also, I want to show a study that Bell and colleagues did, looking particularly at youth of color. They noticed that this discrepancy in parent and adolescent dyads also was illustrated, as I have said, particularly that youth of color – there was more of a significant discrepancy compared to white populations. This has been shown in both – in various populations. This is a consistent finding that we find in the literature.
One of the things that we think about is the response, what happens. I want to note. This is from the Youth Suicide Research Study. I think this is an interesting slide. They looked at racial and ethnic differences about responding actually to the suicide items in that questionnaire. It is done nationally. And we utilize it a lot to understand the rates of suicidal ideation and behaviors and samples.
But there is a racial and ethnic discrepancy in actually who even responds to the suicide attempt question. And particularly, youth of color were more likely to omit answering the suicide attempt question. The data that we get is biased in this way when we do not have this information because if someone does not answer, we do not know. It could be assumed that they did not answer because it is a yes but we do not know that for sure. We are missing this information. As you look at the African American youth, 19 percent omitted answering that question. That is a significant number.
When you think about the reasons for not disclosing that comes up, things that have been identified as stigma. I will be talking a little more about stigma, but that has been permeating throughout this morning in some of the presentations.
And then this could be a perceived low risk of suicide. The youth may perceive, okay, I think about dying sometimes but I am not going to really do it. They think that there is no reason to tell anyone. We get over into where kids may be cognitively and how they understand things in the abstract thinking.
And then there is the fear of consequences of disclosure. We know when a youth tells someone – I am thinking about dying in certain environments, it elicits a set of reactions that they may not want to have to deal with. Particularly in school settings, these kind of comments oftentimes – I know kids come in to see me and they cannot even go back to school until someone says that they are safe to go back to school because they made a comment. These could be reasons why youth are not disclosing information about their suicide risk.
Some of the implications of discrepancies in this lack of disclosure is that we need to interview other people’s circle to gather more information about suicide. I think this has come up. Relying only on a parent is likely not sufficient to really understand everything that may be impacting you particularly when into suicide ideation behavior, previous attempts.
There may be other things that the youth are not disclosing to their parents. Romantic relationships that they may be involved in. As it came up, sexuality, gender identity. And then there could be trauma that they have never told. There could be bullying that they may not have told their parents about. There are things besides suicidal ideation, other topics in which parents may not be aware.
Next, I am going to move on to talk more about the stigma and beliefs. Suicide bereavement is perceived differently, in particular that once a family becomes bereaved by suicide, society can view them more negatively. There is stigma associated. Social isolation that may occur. This is perceived differently than if a family member may have died by natural causes or even an accident. There is research showing this difference.
Also, there is stigma experienced by the survivors. They may feel guilt, a responsibility, shame. This has been identified and these factors come in play. They are having their own internal struggles as well as the community in their surroundings others may be putting on it. They are getting a double whammy.
Survey done and these are with adults, of individuals with a lifetime history of suicidal behavior. This was done through the American Association of Suicidology Listserv. And what they found from these people who completed the survey is that they have perceived stigma from a family member was the highest rate of the stigma that they perceived.
Adults and probably youth, which I think we can extrapolate, are experiencing stigma from their families based on suicidal ideation and behaviors. They may be less likely to disclose it and they may have an impact. And when they talk to families about (inaudible) because of stigma, they may be less likely to admit that suicide could be a possibility, which came up earlier this morning.
Some of the beliefs within the larger community besides the social stigma is there is this othering. Other people may kill themselves. They separate it from that it can happen with anyone and make these differences. There is a desire to rationalize why and we may not know why. It is pretty common to have these stigmas and then also the way the media represents suicidal ideation and behavior and people who are bereaved by suicide. Many times, these are negative connotations that people see. You want to distance yourself from those. And then there is this relationship with psychopathology and people may view severe psychopathology with suicidal ideation behavior, which is not necessarily.
One of the things that came up is that when we asked people about suicide. People are afraid to ask children. But what happens when you – if you ask them about suicide, they believe that youth may put that idea in their mind and say that is a good idea, which there is no evidence to support this. There are concerns about this but most of this evidence with older youth. But there has so far been no evidence that it is the case with younger youth.
One of the things that has been identified – there was a qualitative study done with youth 10 to 21 in emergency departments, Ballard and colleagues. And they asked youth what did they think about universal screening, asking every kid that comes in the emergency room about suicide. And what they found is actually that youth got it. They actually were mostly supportive. Only 10 percent had some opposition. If you looked, having a negative risk that was only in 2 percent of the population. Even though it does exist that there is negative potential that people may perceive it for generally asking about it, it is actually a shift that people recognize that it is more important and important to do.
This is also a meta-analysis of the impact of suicide-related content. What they found – and this has both adult and adolescent samples – is that asking about suicide-related information has no effect. Effect size is zero on someone’s distress level and functioning.
Actually in this meta-analysis, they actually found that they looked at people pre and post before and after – being exposed to suicidal intent that actually adolescents had some benefit. There was one study. I think there is still to me more evidence that is needed but there is clearly no negative impact from this meta-analysis.
I do want to talk about the data for young children. As we have been discussing, there is limited data for children under 10. An example is the NVDRS articles. When they looked at kids, children, and early teens, this is the age group between 5 and 14, only 17 states actually were able to contribute to this database. We are missing data on a lot of states because they are not collecting data on these younger youth.
And also with the CDC WONDER, children under 10 is often excluded as we say because this determination of self-harm intent that has come up.
When we think about intent, Posner identified two ways. The individuals’ own perception of the potential lethality of an act and also the acts of impressive circumstances, action explainable by no other potential intent than intent to die. This has been discussed earlier. Trying to figure out intent, especially for youth and younger youth is complicated and not always very clear. In these cases, any time that – because intent cannot be determined, that is not ruled to suicide and we are missing that data.
One of the things that would identify as a presence of a suicide note has come up. This is probably less likely to happen now and as people were saying. But in one of Arielle’s earlier article where we talked about the children and early teen deaths. Only 7.7 percent of children actually left a suicide note. However, 30 percent of the teens did. There is a discrepancy.
There was another article, which focused only on black youth, and they also found this discrepancy in age. That for older teens, they are more likely to leave a note than a younger youth. There is this limited information that we may have about these younger youth.
I want to go back and talk about that came up about the racial/ethnic misclassification. This has come up as a particular issue. I think particularly this is irrelevant as we think about youth and particularly that in the US, racial and ethnic multi-racial youth are increasing and so how that will even be captured when we know that there are some groups that even identifying race as complicated or misclassified and then we think people may have two and three different race or ethnicities. It is important to gather that. And it is not clear how the family may identify. Is it similar to how a youth may identify their race and ethnicity? But this limits our accuracy in gathering the suicide mortality data.
This is just some research from an article. I noticed the report was shown earlier. But this shows for Native Americans and how they are misclassified. Particularly, many of them are identified as white who are Native American and indigenous. For ethnicity, it is a lot better than with Native American but there are still some discrepancies in misclassification –
I will end here, and I will send it over to Stephen O’Connor, who will be next. Thank you.
STEPHEN O’CONNOR: Good afternoon. I am Stephen O’Connor. I am a program officer here at the NIMH and I am Chief of the Suicide Prevention Research Program in the Division of Services and Interventions Research. I am really excited to have the opportunity to speak and to be part of this workshop.
The diverse groups that are represented in the room here – this is sort of the story of suicide prevention in the United States honestly because in the mid-20th century, there really was no real understanding about the internal experience of suicide. We had mostly sociological research that understood social factors that conferred risk. But in 1949, a young psychologist named Ed Shneidman was at the L.A. VA and he was asked by his service chief to write a note to the spouses of two veterans who had died by suicide.
Ed Shneidman reviewed their medical records and then went to the L.A. County Coroner’s Office and asked for permission to read their suicide notes. He read their suicide notes and he learned a lot more about what was behind their decision that they made. He also discovered that he was in a room full of suicide notes. And therein, he began a journey of understanding the suicidal mind. It was really an intersection between the different groups that are in this room and here we are today benefiting from the work that each of our groups has done and now trying to tackle a new issue but trying to bring out the best from all the work that we do individually and find room for synthesis. I just wanted to share that.
Really quickly just to orient you to the structure of our NIMH Extramural Division, we have different divisions that really span the translational pipeline of the type of science that we fund. Each of these kind of maps on to these different goals that we have in the NIMH strategic plan. Those at the bottom you can see that kind of maps on to different stages of clinical research. In the division that I am in, often the clinical research that we are funding is testing the effectiveness of interventions that have bene proven to work under more controlled conditions. They are efficacious. And now we are trying to take them out into real-world settings and see how effective they are when you study them without the added benefit of a lot of research infrastructure.
This topic around how to help those individuals who have recently lost someone to suicide has come up several times already today. Clearly, everyone knows that this is a gap issue in both the clinical realm as well really in the research realm. There has been some very nice research that has been done but a lot of it is more recent and a lot more needs to be done especially translating the research that helps us identify who might be at risk into actual prevention strategies that can be used.
There has been a few very rigorous studies to look at through the ripple effect of losing someone to suicide. One study that was conducted by Julie Cerel in the State of Kentucky found that for every person who died by suicide, there were 135 people that were exposed. If you do the quick math and you say there were about 48,000 people who died by suicide in the United States in 2021, that means there were about 6.5 million people that were exposed to a suicide loss in that year.
It is important to avoid pathologizing a painful and normative mourning process that people are going through. However, we do know from previous research that parental loss of a child to suicide is a risk factor for suicide death, for depression, anxiety, and for marital breakup. There is a stronger effect for suicide risk when the death occurred during early childhood. Regardless of whether the manner of death is through self-inflicted or if it is an accidental injury, risk is much higher in the first month following a child’s death in both men and women but it remains markedly high for women across the first year. This tells us a little bit about the immediacy of risk after the loss of a child.
We also know that in general, about one out of five youth are exposed to the loss of a peer to suicide. Exposure to a schoolmate’s suicide has been shown to increase risk of suicide ideation and suicide attempt cross sectionally but it also predicted risk for a suicide attempt across two years in youth ages 12 to 15.
Importantly, the risk of that future attempt was not attenuated by the closeness of someone to the person who died. That kind of speaks to the importance of more universal prevention strategies that schools and communities could use to help support youth when there has been a suicide loss.
There is a national strategy for suicide prevention. There is a national action alliance for suicide prevention. As part of that action alliance, there are different workgroups. There is a workgroup around for clinicians. There is a workgroup for people with lived experience, people who have experienced their own suicide thoughts, suicide behaviors. And there is a workgroup for loss survivors as well.
The loss survivors created their own report that can be found on the link at the bottom of this slide. It is chock full of really helpful information. Within that, there is a description of this continuum of effects of suicide exposure. The largest of the circles in the Venn diagram are the suicide exposed. This is everyone who has connection to the deceased or to the death itself, including witnesses.
And then there is a smaller chunk of folks who are personally affected by the suicide death. Now, there is a smaller group of folks that you could describe as the suicide bereaved in the short term. And then there are those folks who could be described as suicide bereaved in the long-term and those are people for whom bereaving becomes a protracted struggle that includes diminishing functioning in important aspects of their life.
It is important to think about ways that we can help everybody represented in this image but then also maybe have tiered or stepped approaches to help those people that are struggling in more unique ways.
Ed Shneidman, who I referred to at the beginning, also created the term postvention. He loved neologisms. That was one he created. I read that he referred to that term postvention at the first meeting of the American Association of Suicidology, which I think took place some time in the mid-1960s.
But ideally, the essence of this is that this is a very specific context where we might be able to help people who might be at elevated risk of suicide. At NIMH, we are often very interested in these unique contexts where we know people are experiencing heightened risk. When people leave acute care centers if they have been in inpatient psychiatry hospitalization that when they leave and they transition back to the community, that is a really high-risk window. The postvention window, as I was showing you the statistics earlier, on the data that shows that within the first year, parents, when they lose a young child, that is a really high-risk period for both parents.
The goal here as described in that National Action Alliance Subcommittee report is to prevent further suicides in the postvention window, to support the bereaved, and to counteract other negative effects of exposure to suicide. Previous presenters, as part of this session – just really blown away by everything that was described. I think a lot of it kind of resonates with discussions we were having earlier and then this idea that risk exists at multiple levels.
If you think about Bronfenbrenner’s model, the Social Ecological Model, potentially in the postvention window, there could be risks that exist within a person, their own psychological processes. There could be risks conferred by the interactions that they are having with another parent, with other family members. There could be risks conferred by the messaging that they are getting from society or their community about what it means to be a loss survivor. We need prevention strategies that address the factors that confer the risk at each of those different levels.
Within that report that the Action Alliance created, there is an emphasis on meeting the needs of the suicide bereaved. One of the things that they highlight is compassionate assistance from first responders. This was published in 2015-2016. Everything I can tell about everything that has been described here today, there is a lot of compassion. That is why we are here.
But I want to read this excerpt because I think it is very powerful.
“Compassionate assistance from first responders. Professionals who are likely to have early contact with the bereaved, including police officers, coroners or medical examiners, emergency department personnel, faith leaders, funeral directors, can have a significant impact for better or worse on the newly bereaved. They can be supportive and compassionate, providing comfort, needed information, and linkage to other resources. Or they can be detached, impersonal, and even hostile to loss survivors, focusing only on the technical requirements of their job rather than on helping the human beings they serve and ameliorating the shock and tragedy of suicide.
The bereaved deserve a consistent response and first responders who are well trained, knowledgeable, and compassionate in their approach to new survivors. In addition, first responders themselves and others exposed to suicide may need assistance with their own grief and trauma reactions that result from their work experience.”
There are excellent examples of resources for families, schools, and communities that can be found in Appendix A from the Survivors of Suicide Loss Task Force.
Here is another image of the concentric circles. I guess not so much a Venn diagram. The task force laid out these different strategic goals. I will just give you a description of what each of these looks like. The universal strategies focus on healthy and empowered individuals. These are interventions on behalf of the entire population. A few examples include integrating and coordinating effective suicide postvention activities across jurisdictions, organizations, and systems through increased communication, collaboration, and capacity building.
For selective strategies, these are interventions on behalf of people who are at risk for negative mental health, for unhealthy outcomes in the aftermath after a suicide. Those might include developing and implementing protocols in communities and across caregiving systems for effectively responding at the scene in the immediate aftermath of suicides and ensuring that people exposed to a suicide receive essential and appropriate information.
For the indicated strategies, these are interventions on behalf of people who are currently experiencing negative outcomes as a result of suicide. Goals there might include ensuring that all support and treatment services delivered to the suicide bereaved are accessible, adequate, consistent, and coordinated across systems of care.
And then underneath all of this in our learning health care system type of way that we love so much at NIMH are increasing knowledge base so increasing surveillance, research, and evaluation investments.
If you were considering a way to try to develop or test a preventive, therapeutic, or services intervention, at NIMH, we would encourage you to consider this idea of experimental therapeutics. All the research that we fund – the clinical trials are based on this idea that not only are people studying the main effects or main outcomes of an intervention but like Dr. Luby was referring to earlier, really discovering the key role of a target mechanism. She was describing the key role of emotion regulation and parental validation that help lead to changes in the distal outcomes of those depressed youth.
You would want to step back here and ask yourself a little bit, do you think that there really – the treatments that exist that are for people that have experienced trauma, that somehow they would not be effective for loss survivors in this context. If that was the case, there would need to be an empirical justification and a good conceptual model and you would say I would like to develop and test one of these preventive or therapeutic interventions that is delivered to an individual family or group.
However, if you thought maybe one of the issues is not so much having to create new treatments but it is about changing systems of care, you would focus more on the bottom of this diagram where there is a services intervention that is really meant to address targets or these are factors that facilitate or impede use of services, delivery of services, and better outcomes. Those would lead to improvements in access, engagement, quality, equity, efficiency. I heard a lot today. The issue is resources.
We face this in primary care too because we know so many people who do end up dying by suicide who have access to health care and a lot of those folks have access to primary care settings. But it is not like we can say a primary care physician needs to become an expert psychotherapist. It is just not realistic. There is the capacity to do that, but to actually fit that into your workflow and being able to deliver that maybe just not realistic. I am kind of hearing some of the same things from the presentations that we heard earlier today.
Let us think about maybe what candidate intervention targets would look like. At the individual level, there has been nice research that indicates that meaning making is an important internal process that is associated with healthier outcomes for parents who are bereaved by the death of a child to suicide.
Meaning making, of course, is highly personal. One way you could think about it – maybe thinking about that from the trauma literature. A lot of times when people experience trauma where we have these schemas, these ways that we think about life, a lot of us grow up thinking that the world is a just place. We have a just world theory. And then when something traumatic happens, if we assimilate what occurred into our schema, we would say good things happen to good people. Bad things happen to bad people. If that is the belief that I hold on to, then that must mean that I am a bad person or in some way, I deserve this.
If you over-accommodate a schema, you would say you know what, the world is just upside down. It does not make any sense. You do not know if good things are going to happen to good people or bad things are going to happen to bad people. Anything can happen at any moment in time and you might become hypervigilant, look over your shoulder. Those types of symptoms are reflected in the DSM criteria for PTSD around changes in cognition, emotion, hypervigilance, and avoidance. But that path towards meaning making kind of means understanding your own personal narrative and how you make sense of something, a painful journey but necessary.
Potentially focusing on grief and depression. Guilt and shame. We heard that mentioned earlier from earlier presentations. And then perceived avoidance from family members. This is a study that came out recently by Westerlund and colleagues. They were focusing on potential targets among mothers who had lost a child to suicide. They wanted to see what conferred risk of suicide ideation among those mothers. Those guilt and shame and perceived avoidance from family members were the two main factors or mechanisms that they identified. And those kind of map on to one of these theories that we think about suicide prevention often, Joiner’s theory, feeling like a burden to other people, and that somehow they would be better off if I was not alive and feeling unbelievably disconnected, as well. Those two factors help create the context for suicidal thoughts can be generated. That makes a lot of sense to me.
At the school level, school connectedness has been associated with reducing risk of suicide. You could break that down into four smaller categories: social affiliation, school belonging, attitudes about school importance, and supportive learning environments. These are things that can be done all the time but they create a more supportive environment so that the school network can do a better job of experiencing the loss of someone to suicide. There is more protection. It is already built in.
But then also, having greater access to evidence-based health care services and school staff competency and self-efficacy. We know around training clinicians that their self-efficacy, their belief, that they can do the treatment well. That goes a long way in terms of their ability to help people.
Candidate intervention targets for community level. Economic support, something to consider. Creating protective environments. That might have to do with safer storage of lethal means. Promoting connectedness in communities, teaching coping and problem-solving skills, identifying and supporting people who are at risk, and safe reporting and messaging about suicide.
Here are a few examples of current NIMH now notices of funding opportunity announcements, NOFOs, that people could consider looking at and they might be able to support research in this area. Pilot effectiveness trials for treatment, preventive, and services interventions. Those involve clinical trials. We also have a NOFO that does not involve clinical trials. And then more exploratory, high-risk, high-reward studies that are often there to help identify what the mechanisms are if you do not feel like the literature is strong enough to support maybe one of the R34s at this point.
Thank you very much.
BRENDAN WEINTRAUB: Now, we are going to move into our discussion. I would like to invite everybody, Jane, Arielle, Joan, everyone to pop up on the session, pop on to camera and we can go ahead and get started.
JANE PEARSON: Great. Fantastic presentations, everybody. I wanted to cover two things very quickly and then we can get into some more of the content. One of the viewers had asked if the 988 crisis line was also good for Spanish-speaking persons. Yes. There is a whole Spanish page if you also want to do the Spanish chat. I wanted to make sure we address that.
The second thing is – Arielle, thank you for bringing up taking care of ourselves, as we are working in these topics. We actually had a great question. They were curious to see if – let us just say any of you all – how do you take care of yourselves? And just do it quickly in the room to get back to the content but might just give some folks some ideas. Maybe just a quick suggestion from each of you.
Stephen, since you were just up, why don’t you go ahead?
STEPHEN O’CONNOR: Exercise, spending time with family, lots of hugs.
JANE PEARSON: Arielle, how about you?
ARIELLE SHEFTALL: I would say the same thing. But also, I love classical music and singing. That is something I try to do as an outlet as well.
JANE PEARSON: That is great. Joan.
JOAN LUBY: I think just focusing on joy and having fun in whatever way you can find it.
JANE PEARSON: Great. Rhonda.
RHONDA BOYD: I will repeat. Spending time with family and friends is important and doing activities. I love reading that is not related to academia. A good mystery is always relaxing.
JANE PEARSON: Thank you. It is really an important topic. I am glad it got brought up. Thank you for the question and Arielle, thanks for bringing it up.
I will let you all ask questions of each other. I have a bunch and then there has been a few up on the Q&A. Let us see if you all have some questions for each other first. You were referencing each other so well that I think you all probably know each other’s literature well.
Why don’t I go to – one of the questions that Rhonda brought up that links to this morning is the ethnicity/racial accuracy and also sexual – when there is multi-racial categories of individuals, how have you all seen that? We are beginning to see it show up in some of the surveys. For those of you looking at this data a lot in terms of reporting rates, what have you seen? What do you think looks helpful? And then maybe we can by the end of the day pull in our earlier panels to think about how NCHS thinks about this, how CDC thinks about it. But what have you seen that you think looks helpful that maybe should be considered in data collection?
RHONDA BOYD: I guess I can start. I just know what – with my research teams as we have – we just had this conversation a couple of weeks ago, thinking that the ‘other’ category is not sufficient. I always encourage to allow people to pick whatever race and ethnicities they want. Pick as – and then we can put together – I am going to push NIH because a lot of times we put things based on what NIH wants, and so that does not allow that very well. I know there is some discussion about that at the level about how we actually capture that data because I think we are missing things and forcing people into boxes that they do not maybe need to be in.
JANE PEARSON: Let me just also throw in there that I think youth in terms of their identity development, which is really important for their mental health, think this way. And if we are not asking them or even talking to them or helping them understand how to put these different pieces together, we are kind of missing it. Arielle or Joan, comments there?
ARIELLE SHEFTALL: I can speak to that just a little bit. I am going to be doing a presentation on minority youth, and one of the requests was to look at youth who identified as multi-racial. And when I looked at the national databases that I use all the time, the data only goes back to 2018. We know that multi-racial groups are the largest growing bodies present in the US. Unfortunately, we just do not have the data to quite understand the trends that we are seeing and how suicidal behavior risk factors, protective factors might look differently in youth that identify as multi-racial versus those that do not.
JANE PEARSON: Anybody else? Jan in the room has a question.
JAN GORNIAK: My question – like I said, I am really enjoying all the information that is being shared with us today and even the title of this meeting, child and youth suicide. Now, I am just wondering what ages capture child and youth because I see youth and I see child. Are there age groups like are we going to call them school age, toddlers, tweens? Is there a distinction between the ages of the kids that we are talking about?
JANE PEARSON: Really, a simple question. I think we struggle with that because we have to put it in the announcements. We will talk about preteens. Is that 12-year-olds, 11 year olds? We try to define it. And depending on what you are focused on, you could talk about grades in school, or if you are looking at a system of care, the pediatrics could expand to young adults. It is a huge problem and I think the only thing we can do is ask people to really define it when they are using it.
But if this had implications for guidelines for NAME, I think there would have to be some explanation around this and I think Joan’s work speaks to the cognitive status being even more important for young children and their emotional intelligence or whatever you want to call it, to describe this. That would be the first disclaimer in terms of how to do this.
And it also speaks to why an age cutoff is probably not a good idea for a policy for defining some groups in years per se. But I will open it up to the rest.
JOAN LUBY: We think there’s a meaningful difference as to the 8 and 9 and up and – it is under 8 and 9 – that is a much newer domain and developmentally very different.
JAN GORNIAK: The reason I asked that question was because I was wondering when we started talking about prevention strategies. I know there is overlap. There is a difference between a 6 and an 8-year-old but then that 7 year old is – you know what I mean. I was just wondering how we classify this. Thank you.
JANE PEARSON: Are there other questions from the room? I see one question about – a lot of questions for Joan because I think she covered both descriptive and – and we got something that works so that is quite exciting and it got a lot of responses on the chat on the questions.
Where should we start? One thing would be in terms of screening and actually Rhonda could also help with this too in terms of are there some screening forms that seem to work better for kids and probably in terms of gathering more accurate information. How do those possibly vary by school or by practice just generally like what are some better instruments that you are aware of?
RHONDA BOYD: One of the – for depression, we use at the Children’s Hospital of Philadelphia the PHQ-9 frequently. The American Academy of Pediatrics recommends depression screening for kids 12 and older. How it looks for younger kids – this is the thing that we always struggle with. The PHQ-9 is sort of acceptable across a lot of different environments and it has a suicide item per se, but there is controversy in how well that will pick up if someone has a suicide risk and what you do to follow up with questions.
We also use the Columbia Suicide Rating Scale in our outpatient clinic. There are forms that are for lay people that can be utilized. I know in our pediatric center the ASQ you as being used also.
JANE PEARSON: ASQ. Joan, related to what you were referring to in the open stem questions puppets, people are also wondering if books or videos could be helpful, maybe more in a therapeutic setting to get a better sense of what kids are thinking. Do you have experience with that?
JOAN LUBY: Just to get back to the screener question, to screen for depression of course you can use the CDI and that goes down to about age seven. We have also developed something called the Preschool Feelings Checklist, which is pretty well validated down to age 3, which is a very brief ten-item, two-minute checklist, easily available on our website. That has a suicidality item on it and now my colleague, Laura Hennefield, has elaborated on that and has produced an early childhood screener, which she is now testing. I do think we will have a screener.
Utilizing books and videos is always a good intervention tool. I do not think it is a very pragmatic screening tool because you just – when you screen, you want it to be very quick and feasible and cost effective. And if it requires a human to sit down – that is just not going to work. I think screening is better done with a caregiver.
JANE PEARSON: That makes sense and actually the question was more about a therapeutic part of it but it was almost if the kid was in a clinic having watched that then following with a screen. It is sort of a prompt or maybe making it okay to talk about I think was more of the flavor of that.
John in the room has a question.
JOHN BLOSNICH: Maybe this is for Joan and maybe it bridges over to Stephen as well. Thinking about siblings of a young person – I do not know, Joan, if siblings were – some of your analyses – the role around siblings or how do we think about services for treatment or whatever sort of prevention or postvention we do not just around the parents but around if the kid had any siblings. Is there any research on that?
JOAN LUBY: A really good question. I don’t think there is a lot of research on that. I do think some of these parent-child interventions radiate to the siblings because then the parent – if the parenting style changes then it impacts the siblings. I am not aware of a lot of studies that looked at suicidality in siblings.
STEPHEN O’CONNOR: But I do know that there are resources available via AFSP.
JANE PEARSON: We are losing your sound, Stephen.
STEPHEN O’CONNOR: There are resources available but I do not think that the research has been done to really understand --
JANE PEARSON: It is a great point. Margaret, you have your hand up. Arielle, did you want to answer? Go ahead.
ARIELLE SHEFTALL: I was just going to say really briefly that a paper that we did where we looked at the narratives – youth who had died by suicide – these younger youth. One of the actual things that we noticed is that they had an argument with a sibling or they have told a sibling prior to their death. I think it would be really important for us to consider how do we incorporate siblings into interventions for prevention programming so that we are equipping siblings as well as the parents so something to consider for future interventions and prevention programming.
JANE PEARSON: Great. Thank you for that. And before I get to Margy, there is a question about ACEs. Is that a way to go in terms of the different questions that are now out around experiencing that or should you be asking a little bit more targeted questions, I think, or ones that might get more towards even resilience?
The thing about siblings is kind of interesting because let us say there has been experience of abuse and neglect. If you have a sibling who is also aware of that, that can really change the environment here in terms of before a death, certainly intervention, but can also certainly affect how you are going to be dealing with a bereavement, I think. It kind of reverberates. I just want to connect those two.
Let us address the ACEs thing first in terms of understanding the broader background and maybe that would feed into NVDRS and we have not even heard about our child death review folks and all the data they are collecting. Let us see what people have to say about that right now in terms of what – again, maybe what instruments you are finding that are useful.
JOAN LUBY: I think the ACEs is – obviously, it is a robust risk factor for lots of very bad outcomes. But I do not think it is going to be that useful as specific risk factor for suicidality– surely, the experience of trauma, bullying, rejection, invalidation, but that is not – that is different than ACEs in and of itself. I do not know that ACEs is the thing to zero in on.
JANE PEARSON: Great. Anybody else?
RHONDA BOYD: I think there are things in ACEs that are pretty common that we do not know about, particularly that they have been modifying to get information about ethnically-diverse youth. Is there community violence? Is there racism discrimination? Bullying does not capture that necessarily. The incarceration of a parent and what that separation means, things that disproportionately happen to youth of color. I do understand how some of those things could be important to understand the context. I am not sure if they come out in these interviews that were talked about earlier. There are things people may not be that open to discuss when you are meeting with the family and those are delicate topics to discuss.
JANE PEARSON: There was a checklist that was brought up earlier that I was like could I see that closer and that might be a way to start looking at some of these concepts we are talking about and are they ways of getting to that in an interview. Again, there are challenges with this.
I am going to go to Margy and then Jan.
MARGARET WARNER: Thanks, Jane. You did set me up again for the question I had, which was one of the previous speakers was talking about screening tools and diagnostic protocols and different things they are using with youth when they are a lot. But I am wondering if any of those protocols or screening tools could be adapted to post-mortem settings. And then Jane brought up the checklist that – I think it was – that Joyce showed where it was the questions that were looked at and responded to in the case management system and how – are there options for bringing some of that state-of-the-art diagnostic tool to the post-mortem setting.
And the concept of psychological autopsy sometimes comes up, which given what Lauri said about in the moment what a death investigator, somebody in an office of a medical examiner maybe actually – or a coroner may actually be able to do. How concise can we get those questions down so we make sure we get at the heart of the issues?
JOAN LUBY: I think that is a great point and really great idea and probably very under explored. I would think the issue would be that psychological post-mortem would need to be conducted at a certain point in the grieving process when the caregiver could give an effective answer. But I think that could be really important --
JANE PEARSON: I think we are going to hear a little bit about some folks doing that.
Jan, go ahead.
JAN GORNIAK: I just had a comment when someone was asking about siblings. So when I was in Ohio, any time we had a suicide of a young person, we would call children’s services. They also – especially if you knew there were siblings in the household. People hear children’s services – they are like “oh no. I did not do anything”. But they have a lot of great resources. Obviously, they also sit on the child death reviews. They hold a lot of information. Any research or anything like that, you probably could reach out to children’s services with some information.
But to help those siblings, like you said, there might be more information that comes out with the support from children’s services. That was just a point that I wanted to make.
JANE PEARSON: When Stephen was going through his types of interventions that could happen and I was thinking about everybody saying we need more resources, and we need more resources in suicide prevention at all levels. But are there some services we could leverage and use in a way that do really help make those connections is a big issue.
I do not know, Stephen, if you have any more thoughts about how you could leverage how the ME/coroner community could see – do you think telehealth would even help just to make that connection saying we can’t get there right now but we are going to be talking to you in the next few days and we will be there to – we will help you walk through this. Even that sort of helps saying it is coming. We are here to talk or call us back. I do not know. Any thoughts about that?
STEPHEN O’CONNOR: I think that could be really beneficial. We know that when we interact with people who have just survived serious suicide attempts that there is often a window. There is an openness and an opportunity. Over time that window does close because people kind of get swept up into everything else that happens after you survive an attempt.
I would imagine – there is the potential to do very unique and beneficial things in that near term after someone has experienced a loss. I guess we just need to understand the bumpers. We do not want to open people up and leave them hanging. It is going to make them even more desperate. But really, I think helping provide just basic facts, good information, and then not over promising but at the same time, being able to deliver. When 988 was expanded, it was really important to have a sufficient workforce that could take on the increase in utilization. That has been really important.
I think what we heard earlier today about being adequately resourced is really important so that people can follow through maybe on the recommendations that they are making because it is not going to be helpful if someone gets their hopes up only to have it dashed because well, I was going to do this but I cannot find anybody to work with me or help me. Then that is just going to make someone feel maybe potentially hopeless or not cared about.
JANE PEARSON: Great point.
Holly in the room has a question or a comment.
HOLLY WILCOX: Hi everyone. I was going to speak about this tomorrow but we, in Maryland, have a partnership with our medical center’s office, the office of the chief medical examiner, and we send – we have been conducting psychological autopsy interviews with family members, siblings, and parents of youth who have died by suicide. One thing that happens that is interesting is we are going over this interview trying gather short-term risk factors but I think there is a great deal of benefit that the families derive from being able to talk through a lot of their experiences.
Actually, one of the psychiatrists I work closely with who does these interviews got a note from a dad saying he helped my son whose sibling died by suicide grieve and helped him a lot in the grieving process. Sometimes these things that are set up as research has these interesting, beneficial impacts and it makes a lot of practical sense that it would really give that person the space to be able to talk through a lot of their issues and things that they have been grappling with in their grieving process.
But I just thought I would mention it but more tomorrow on that.
JANE PEARSON: Other people in the room, any comments? We are building in some pieces of the research cycle here. We have some pieces on better questions to ask, valid questions, the challenges in asking the questions. We have a flavor of some intervention work especially with the preschoolers, which is really rare in terms of trying to find very targeted information. A lot of the universal interventions do not identify this really high risk group that Joan has been talking about and that is important.
It is interesting to think about how similar it is for DBT-like skills for teens and adults that you are building. You are starting really early and probably teaching the parents those skills along the same way. That is really interesting and how could that be done even broader.
Somebody had a question about – some editorial about schools getting all this money but not knowing what to do with them. Building curriculum, testing it. We know that is really a big lift for schools. But we are seeing, I think – or hearing a lot here in terms of what could be done to generally support kids, probably staff, and even think through the postvention piece.
Stephen was pointing to a lot of resources at the National Action Alliance. SAMHSA also has a toolkit for schools. There are pieces of information like that that could be helpful probably to the MEs to let them know and hopefully the state suicide prevention offices know to get these postvention materials out to know to screen kids in schools when there has been an event like that.
We are going to hear more about the CDC expertise in doing Epi-Aids and looking at suicide clusters. I think that will also be really helpful to hear more about.
Have I missed any questions. We could actually end a few minutes early. Give you guys a break. It has been an intense day so far. Anything else from anyone? Great panel. Thanks, everyone.
BRENDAN WEINTRAUB: Welcome back, everybody. As we have done for the past few breaks, we are going to start off with a quick Mentimeter. Again, you will be using that same information for it, going to www.menti.com, using that code 31733. This is going to be a word cloud. What is one word you would use to describe the current state of data related to youth suicide? I will note that you can say something up to three times. If you see something on the screen that you would like to further emphasize, feel free to write that same word. It will get bigger. It looks like limited, underreporting, improving. Again, we are just going to leave this up for folks to take a look at as we are going. But I am going to pass it over to Dr. Karin Mack.
Agenda Item: Overview of Coordinated (Post Suicide Determination) Data Collection Approaches
KARIN MACK: Good afternoon, everybody. Like Marcus and Margy, it is just great to see everybody in the room after all the planning to pull this together to see the rich discussions that are happening. It is just amazing.
I would really like to thank NIH for hosting us. I would like to really honestly thank the speakers for being willing to take the time to travel here but to also to take time out from the important work you are doing to help contribute to this content. It is very important and we appreciate each of you for doing that. Thank you very much.
We have another round of fantastic speakers for this afternoon, and I am sure will lead to thoughtful discussions. The first two speakers are going to describe some robust data systems that help give us insights for prevention. Our third speaker is going to give us some examples of helping a community understand and describe the situation with suicide cluster and then what steps communities can take to move forward.
The bios are on the website. Our first speaker will be Abby Collier, who is Director of the National Center for Fatality Review and Prevention. The second speaker will be Dr. Janet Blair, the team lead for the Mortality Surveillance Team and the Surveillance Branch of the Division of Violence Prevention at CDC’s Injury Center. And then finally, we will conclude with Dr. Steven Sumner, who is the senior advisor for Data Science and Innovation at CDC’s Injury Center. Thank you.
ABBY COLLIER: Good afternoon, everyone. I am Abby Collier, and I am the Director of the National Center for Fatality Review and Prevention. I want to start this presentation off by saying we could not do this work without our partners. And while our national and federal partners are vital to this work, I really want to thank our state and local fatality review teams. We could not do this work without them. Literally could not do it. They do the reviews. They do the data collection. And they are the most amazing, creative, and passionate group of people you will ever work with.
I want to acknowledge the importance of our partnership with HRSA. They are our primary funder. However, everything I am going to share are my own opinions.
A little bit about the national center. We are celebrating 20 years of continuous funding to support fatality review teams. When I am asked to describe what we do, I say in a nutshell, our job is to say yes and make the work of fatality review easier at the state and local level. Our work divides into two primary buckets: programmatic technical assistance and data technical assistance. Across both of those buckets, we do a wide variety of technical assistance from site visits to coaching, presentations, library of written and digital resources. I have been doing this work about 15 years, and in that time, we have never told someone no for a resource when they have come to us needing something. It is not a challenge. I am not asking anyone to get super creative. But we do really try to meet the needs of our state and local teams.
There is a national network of child death review teams. There are programs in all 50 states and the District of Columbia. There is a few within tribes as well. We have more than 1500 local teams throughout the US. There are two main structures for child death review programs. The first is reviews happening at a statewide level and the second is reviews happening at the local level. We have a handful of states that conduct reviews at the state or local level, but they are triaged in what can be a kind of complicated way.
There is a wide range of legislation that supports child death review and some states have no legislation so everything from mandating to permitting to nothing. The funding for child death reviews is also pretty broad. We have programs that get no explicit funding. They are doing it really borrowing from other programs, sharing FTEs, all the way up to a state that gets about $1.2 million in funding to support their child death review program.
Most of our programs are funded by a combination of state and federal grants. Because of that wide range of funding, we also have a very wide range of FTEs. We have some states where coordinating their child death review program is literally 5 percent of somebody’s job all the way up to programs that have 8 to 10 staff. There is this huge variability, and that variability will be a common theme through everything I talk about today.
One thing I want to point out for this group is that while every state sets their own case selection criteria, we cannot mandate anything. We ask nicely and before the pandemic, we brought cookies. Now, we just ask nicely. Ninety-six percent of states are reviewing their suicides. While we cannot dictate what they review, suicides are really at the top of the list of things they are looking at.
It is common in our office for us to say if you have seen one child death review program, you have seen one child death review program.
I know many of you participate in child death review but I am going to do just a quick level set. What you see up on the screen is a very simplified process of what child death review looks like. We think of it at the national center in three key steps. The first is to tell the story. The second is to collect the data. And the third is to catalyze prevention. As I described on the previous slides, there is so much variability in child death review. If your state or local team is not doing it this way, it does not mean you are doing it wrong. We kind of say as long as you are working towards the goal of prevention and you are not breaking any laws however you get there, it is totally fine.
In that first step of telling the story, team members bring their records to the review meeting and verbally share what they know about the child, the family, the community and circumstances of the death. I think of this as putting together a puzzle. Everybody brings their own pieces and we put it together. Sometimes it is a very clear picture and we get from A to Z. Other times we have a lot of holes or missing pieces.
After we have discussed the case, we do our data collection and entry followed by prevention work. And that third step of catalyzing prevention is the reason that we all stay around in this work. I use the word catalyze very intentionally. We do not expect our child death review teams to be the doers of the prevention work but they need to be good stewards of the data and share it with others who can do the prevention work.
I am going to transition now to talk a little bit more about the data that is collected through child death review. But before I jump into the data system in detail, I want to talk about the reach of the data system. We like really long names. The data system is called the National Fatality Review Case Reporting System. We do not have anything for short because it is just not how we operate.
It is currently being used in 47 states. Our center supports two review processes, child death review, which is what we are talking about today, and then fetal and infant mortality review, which looks at stillbirths until the first birthday. What you see up there are the differentiation of states using it for child death review, which are orange, states using it for child death review and FIMR, which are blue. And then we have a handful of states and the District of Columbia who are not yet using the case reporting system. We are going to get there. We are pretty close.
Digging into the case reporting system in a more substantial way, it is a free, web-based tool available for child death review and FIMR teams. It was established in 2005 for child death review and contains more than 250 incidents of child death.
The goal of the case reporting system is to collect a wide variety of data from a large number of sources. It is intended to collect key demographics, risk factors, protective factors, and team deliberations.
The case reporting system is focused on how systems function and documenting individual characteristics with that system functionality. The goal of the case reporting system is never to place blame or identify someone at fault.
In addition to the 2600 variables in the case reporting system, states can add a limited number of state-specific questions. It is hard to imagine that we have – that there are things missing in the case reporting system but there are. We have states that add questions to it. In addition to being a data entry system, there is quite a bit of utility built into the case reporting system for states to get their data back out. All of their data can be downloaded. It can also be used to generate standardized reports. Three clicks of a button. States can get real-time access to their data that is in the case reporting system. We also have a data visualization functionality based on CDC’s WONDER, not nearly as sophisticated but we are going to get there, as well as a Tableau environment for our fatality review teams.
There are certain sections of the case reporting system that are available regardless of cause of death. Most of the case reporting system is organized by cause of death. As you can imagine, with 2600 variables and 26 pages of data, it is a lot. We provide the skip patterns in order to reduce the data entry burden by cause of death. However, again, there are some exceptions and suicide is an exception. If a case is being entered and the manner of death is listed as suicide, this automatically appears. However, about three years ago, we added the ability for teams to say I think this child might have intended to harm themselves. It is not ruled a suicide on the death certificate, but it is in that gray area, which many of our forensic pathology partners talked about this morning. They can check that box and get to the suicide risk factors so that they are able to capture those. Other sections that operate like this include child abuse and neglect, infant safe sleep. Really those things that kind of span cause and manner of death.
With so much data available in the case reporting system, I wanted to highlight a couple of things that this group might be uniquely interested in. You can see we are asking a fairly significant number of questions around mental health history, firearm storage, and substance use. We asked these questions not only about the child but also about the parents or caregivers involved.
One of the sections that we are maybe most excited about is a newer section called life stressors. This section was actually born out of a revision of the suicide section four years ago. We were updating it to match the latest science at the time and we kept coming up with this list of things we wanted to know about children who died from suicide. And we really took a step back and said wait. We want to know these about every kid. We want to know what their context in their community looked like, what were their relationship stressors both in their home and in their school. We have developed this ever-growing list of life stressors that capture these contextual components.
These questions are hard for fatality review teams to answer. They are very subjective. But part of the reason we added these was to start a conversation. We are not always talking about social determinants of health, which we know have a huge impact. We put it in the data form. We provide them with a lot of tools and we are hopeful that we are making progress. But we think this will be where we start to see a lot of the interesting information emerge from child death review teams is through the collection of this information.
I am going to pivot a little bit to talk about how child death review or a surveillance process can be useful and nimble. You have heard a lot of data today about how children die from suicide and some really good population-based data with some specific studies. I am going to highlight what we learned from child death reviews. We have navigated the COVID-19 pandemic for the last couple of years.
A great thing about our data system is that it is ever growing and changing. It is very nimble. We were able to add questions related to COVID-19 within about six weeks of initial shutdown in 2020, and then we have revised them. I think I have lost track of how many times. I am sure it is in the code book but four or five times since then. It really demonstrates how these kinds of processes can be very useful at the local and state level for identifying a literal public health crisis.
One of the things we added to the case reporting system was asking our teams to identify the relationship between the pandemic and the child’s death. We asked them to look at it sort of on a spectrum from the very clearcut the child died of COVID, all the way down to what might be an indirect relationship. And then we included it again because we have a FIMR component or that fetal and infant component. We also wanted to track where a birthing parent contracted COVID during their pregnancy.
And then we also left some space to say there was no impact. COVID had no impact on this child’s death. This child would have died whether or not the pandemic was happening. Again, these are subjective data but they have been very interesting to watch our teams navigate over the last three years.
The data I am going to share in the next few slides are from the COVID-19 pandemic. We maintain a robust researcher dataset. It is available for researchers who want to use it. We have about 200,000 deaths in that researcher dataset. Of those deaths, 6 percent were youth suicides and of that, the most frequent means was intentional asphyxia. That aligns with everything we heard already today from our speakers.
The data that I am going to share from this slide forward is not from that researcher dataset. I point this out because the data that I am going to share here looks a little bit different. Our staff at the center can use data a little tiny bit more freely than we can let our researchers use it. And when I say a little tiny bit, I really mean a little tiny bit.
The data that you are looking at here looks at deaths occurring between January 1 of 2020 to March 31 of ’23. There were 29,711 deaths entered into the case reporting system during that time. About 80 percent of those were marked complete. We have of those cases, check box at the end of the form. Some people use it. Some people do not. We never lock a case. Someone can go back and change data entry at any point in time. There are pluses and minuses to this system. I am happy to debate it at any point in time.
COVID-19 was the primary cause of death for 181 kids in this dataset and COVID-19 impacted just over 1200 deaths in some capacity. I also want to point that about 40 percent, 38 percent of these deaths occurred in 2020.
It is not surprising to me that we see 2020 data a little overrepresented. Our teams run 12ish months behind on data collection and entry. And the pandemic had a very negative impact on CDR data collection and entry.
When we look across that spectrum that I showed you a few slides ago, looking at the relationship between the child’s death and the pandemic, you can see that we had – 13 percent where the birthing parent contracted COVID-19, 15 percent were diagnosed or suspected at autopsy. It is important to note that there was no testing available early in the pandemic at autopsy. That is why that is suspected there. It was not a possibility.
But 42 percent of the deaths in that COVID-impacted group are indirectly related to COVID. What are those cases? Kids who were not able to obtain timely substance use treatment, a delay in seeking care. A kid needed emergency medical care. The family was hesitant to go into the emergency room. We saw a number of cases where COVID-19 led to a family having a home birth and them having a terrible outcome.
Here is what we found. You can read the specific numbers on the slide. But it seems that identified children as having a COVID-impacted suicide, those kids experienced more distress than other suicides happening at the same time. While we had a wide number of children who died by suicide, only a portion of those the team felt like they were COVID impacted. But those kids had more distress. We saw them having a more frequent death of a peer, friend, or family member, a higher presentation of depression and anxiety. They had more life stressors, which I will talk about on the next screen. But what is interesting to me and we have not figured this out quite yet is we had a decrease in missing and unknown data in those COVID-impacted suicides. Interestingly enough, teams were able to get more complete data about those kids when our overall data quality decreased during the same time. It is a little bit of a quandary but at least we know that the data that we are talking about right there is fairly accurate.
One of the things that we are always looking at is that new life stressor section. Suicide was the second most common manner of death in that group of COVID-impacted deaths. Natural deaths were the most common manner of death.
When we looked at all of the deaths that occurred from January 1 of 2022 to March 31 of this year, we found that on average those kids had about 1.7 of those life stressors. There are 60 of them. On average, those kids had 1.7.
When we looked at all suicides occurring in that same time period, January 1 of 2020 to March 31 of this year, those kids had 3.2 life stressors so just about double. But then when we looked at only those deaths that teams felt were COVID impacted, we found that those kids had 4.5 life stressor present.
As we look at this over these different groups of kids, we can see that these kids were – a child death review team felt that the death was COVID impacted. They really looked different in the review process.
When we looked at what were the common combinations or common life stressors that we saw, you can see up here relationship challenges at the top. Technology and transition. In that transition bucket, we are looking at things like formal transitions, in and out of child welfare, in and out of the school year. But we are also looking at informal transition. Were they living with a family member or a friend? Are they moving around in that way?
The most common things we saw in other actually belonged in another category, but we can talk about why we dislike the category of ‘other’ later. But we found things like criminal charges, mental health needs. A lot of virtual schooling showed up in that other category instead of school maybe where it belongs.
I hope this data highlights the unique lens that you can get from child death review. At the national center, we are restricted in how we use the data. I cannot show you any cool maps or geocoding or any of that but states can do all of that with their own data and we are always happy to support that.
I want to end the presentation by saying about 18 months ago, the national center with support from our state and local programs identified five recommendations that we felt would improve our review of suicides and ultimately impact our prevention work in a positive way.
I will just highlight too that we have already talked about today – the first is consistent access to records. As we heard this morning, access to records is vital. It is variable. Some states have very comprehensive legislation that gives them anything they want with a subpoena. We have other states that are begging. It is really broad, and it is often based on relationships. Even if there is great legislation, this is still very people-centered work and that is challenging.
The other one I want to highlight is just standards for death scene investigation related to suicide. We really approach this work that everybody is doing the best they can but if I put myself in the shoes of a death scene investigator who maybe sees one or two pediatric suicides every couple of years, they are not common. It is emotional. I cannot even imagine the challenge of going to a scene and trying to remember the millions of things that they have to collect while navigating all of the emotional dynamics of that. We always support more consistency and training across many causes of death but particularly with youth suicide. The full report is available on our website. If anyone wants it, I am happy to share the link.
I will just end by saying child death review teams while our data are not population based and they can be flawed, they do pair very nicely with some of the population-based data that we have talked about already today. While we cannot give you an exact incidence of something and if you use the word rate, our team gets a little twitchy, we can help you explain some interesting descriptives around what you see happening either in your state or your community or at a national level.
I did mention that our data are available to researchers. I am happy to share that we are focusing a supplement of pediatrics on data from child death review teams as a way of highlighting this important work and the important relationship we have with pediatricians. Again, I am happy to share more on that. Thank you.
JANET BLAIR: Good afternoon, everyone. First, I would just like to say it is a real honor and privilege for me to be presenting at this meeting today. I would first like to give a land acknowledgement and acknowledge the Muscogee Nation and the Cherokee Nation whose indigenous lands house the CDC and ATSDR Atlanta campuses. I would also like to acknowledge the 574 federally recognized tribes whose lands house the state, local, tribal, and territorial jurisdictions that we serve at CDC.
I would also just like to give a shout out and thank you to the NVDRS recipients out in the states, the vital registrars, the coroner medical examiner, and LE partners who work on NVDRS as well.
I am going to be giving you all an overview today of the National Violent Death Reporting System or NVDRS. What do we know about suicide? We know that the number of suicides just tells part of the story and suicide can have lasting harmful effects on individuals, families, communities, and society of the people who are left to pick up the pieces. It is a complex problem, as you have heard today with many contributing factors. But we know that suicide is preventable and that information is needed for prevention.
Just to give you a little bit of history about NVDRS. In 1999, the Institute of Medicine issued a report calling for a national fatal intentional injury surveillance system and the National Violent Injury Statistics System was piloted at 12 sites, mostly universities.
And the first appropriation for NVDRS came in 2002 from Congress. And in 2003, the system started data collection in six states. And from 2004 to 2018, NVDRS expanded to all 50 states, District of Columbia, and Puerto Rico. This is a graphic here just showing you the timeline of NVDRS and how the program has expanded over time. This is the NVDRS map displayed a little bit differently just showing you the system by the year data collection started and it is color coded.
What is NVDRS? It is ongoing surveillance to monitor violent deaths and it links data from three required sources: death certificates, coroner/medical examiner or C/ME reports, and law enforcement reports. This information is collected at the state, county, and city level and it is compiled at the national level.
What is an NVDRS case? In other words, what makes it into the NVDRS system? The system collects suicide, homicide, deaths of undetermined intent, legal intervention deaths, excluding executions, and unintentional firearm deaths. It collects information on multiple related deaths of the above types that occur within the same 24 hours and are captured in one incident so things like murder, suicides that may occur within the same 24-hour period.
The data are collected by the VDRS recipients through partnerships and again partnerships with vital registrars, coroners, medical examiners and law enforcement. This information is entered into a web-based application, and I want to reiterate that no personally identifying information is collected in the system so no names, no names of agencies, facilities, law enforcement officers or anything like that, no addresses. None of that is collected in the system.
Trained abstractors enter the data according to CDC guidance and the data provides information for prevention. I will talk about that a little bit later.
NVDRS is unique. It tells the who, what, when, where, why, and how. It is comprehensive and includes all age groups and it combines the sources that I have talked about to get the full picture so details about the incident, decedent information and circumstance information.
As I mentioned, the system uses a multiple data source approach with death certificates, law enforcement reports, and coroner/medical examiner reports. There are over 600 data elements that make up the system and two narratives are written by the abstractor that describe the incident. One is based on the law enforcement report and the other is based on the coroner/medical examiner report. And these narratives can be thought of as short stories that describe the incident in a few paragraphs.
When information is collected, injury characteristics so things like where the violent death occurs, demographic information, age, sex, race, ethnicity, circumstances, which are events that preceded or were known to contribute to the violent death, mental health diagnoses, as well as toxicology or drugs that were in the decedent’s system.
Circumstances that are collected in NVDRS are noted here. These are the ones that are collected for deaths that are suicides or undetermined intent. You can see things like leaving a suicide note, history of suicide attempts, history of suicidal thoughts or plans, school problems, suicide of a friend or a family member, disaster exposure, which also includes the pandemic, anniversary of a traumatic event, as well as who the victim disclosed to if they disclosed to die by suicide and who they disclosed that information to, whether it was a friend or a health care provider or a parent.
This is a graphic showing our typical data flow for NVDRS. When you take a look on the left, a case is initiated when a first source of information about that case is entered into the system after it is received. It is usually the death certificate. And that information is entered by the state VDRS abstractor into the NVDRS web-based system that I discussed.
States can download their data any time. CDC can also download state data at any time. I just wanted to also mention that the state only has access to their specific state’s data.
I am going to present data later on in this talk for 2020. You may be saying, wait a minute. It is 2023 so why are you presenting 2020 data. That is the most recent year available. Because NVDRS uses this multiple source approach again with coroner/medical examiner reports, law enforcement reports, it takes a little bit longer to get that information particularly if the case is being investigated. What that means is that we give our recipients 16 months to collect data on any of those cases whether the death occurred on January 1 of that year or December 31 of that year. We have our close out occurring at the end of the 16 months.
For example, for the 2020 data year, close out occurred in April of 2022. We will be closing out our 2021 data in May of this year. Again, data close out occurs when we get all available information from all source documents that are abstracted into the system. Like the surveillance system that Abby talked about, the abstractor can go in and can update the data at any time. Even if we have had a data close out, the abstractor can go in and add additional information after it is received.
I do also want to note that law enforcement is usually the last piece received and particularly in the case of homicides, particularly those homicides that are still being adjudicated.
What can NVDRS tell us about youth suicide? Again, I am presenting data from 2020 from 48 states, the District of Columbia, and Puerto Rico. It also includes data from 2 states that are not yet statewide so 35 counties from California and 4 counties from Texas. I talked about the time lag.
As you can see, most decedents were male. I am limiting this analysis to ages 10 through 17 so you can see the age distribution here. Most decedents were white, non-Hispanic. Most suicides occurred at a house, apartment, including the driveway, porch, or a yard. Here, we have the breakdown of suicide by the method of injury, weapon used. Pretty even with firearms and hanging/strangulation/suffocation.
Some of the circumstances that were noted were things like a current mental health problem, ever being treated for a mental health problem. As you can see in about a third of cases, the person left a suicide note, which includes suicide texts, history of suicidal thoughts or plans. About a third had a current depressed mood or were currently being treated for mental illness. And almost a third had a crisis in the preceding or upcoming two weeks.
Family relationship problems, school problems were noted as well as the intent to die by suicide. Arguments or conflicts were also common as well as the history of suicide attempts, and intimate partner problems as well.
Some of the limitations of NVDRS. As I mentioned, the data are not yet nationally representative, and those larger states are ramping up to statewide coverage. Abstractors are limited to source document information. Medical and mental health information may be incomplete, and I know others have discussed the challenges of getting that information. And the information that we get is dependent upon the knowledge of the informant so the next of kin may not know when the person is asked certain questions about the decedent. And also toxicology testing can vary by jurisdiction. Not every decedent receives toxicology testing.
What can NVDRS do? It can uncover timely topics and emerging issues so things like the use of sodium nitrate. We are starting to see that in our database as well. It can allow for examination of specific risk factors. It can reveal important variations and patterns. It can lead to a better understanding of circumstances contributing to suicide and it can also help to inform, guide, and evaluate prevention efforts and ultimately help to create safer and healthier communities.
I am going to talk about how the data are used for action and this list is not exhaustive because we have a lot of VDRS partners who are using the data for prevention. I am just going to give you all a few examples just for a flavor of what is happening out in the field.
In California, they are using the data to support development of a budget change proposal to address increasing rates of youth suicide and it is being done through the California Youth Behavioral Health Initiative. The programs include extensive outreach and media campaign and a crisis response effort to offer community resources and support following a suicide attempt or a suicide death.
In New Jersey, the data were used to support a successful New Jersey Health Department and Garrett Lee Smith Grant applications to target youth suicide prevention. And it supports New Jersey’s Ready to Stand United against Youth Suicide, which is a comprehensive statewide collaborative.
In Nevada, the data were analyzed to see intervention points with youth suicides and as a result, Nevada established Project AWARE, which is a school-based mental health and suicide prevention program that works with the Nevada Department of Education to provide school personnel training, including how to respond to mental health challenges.
NVDRS works with several partners. We work with these national organizations listed here. And these partnerships help to enhance access and availability of quality data: the International Association of Chiefs of Police, the National Sheriffs Association, the American Public Health Association, NAPHSIS, which is the Association of Vital Registrars, as well as the National Association of Medical Examiners.
We have something called WISQARS, which is a Web-based Injury Statistics Query and Reporting System, which is an interactive, online database that is free and available to the public. It has data from 2003 to 2020. We also have an annual NVDRS Surveillance Summary that comes out, which is a comprehensive overview by death manner. The latest NVDRS Surveillance Survey is scheduled – it is in production right now and it will probably be released in May.
We also have data available for researchers who are interested in analyzing the data. It is called the Restricted Access Database or RAD. We have eligibility requirements. We can put the data in different formats depending upon what the researcher is most comfortable with. There is detailed documentation to go along with it in a RAD Users Guide. We have shortened our response time to inquiries to one to two days. We used to review applications quarterly. Now, we review them every two to three weeks.
To learn more, you can visit the NVDRS website. You can do a Google search. Just put in CDC NVDRS and it will take you to that website and you will be able to find out more information about the system.
I would just like to acknowledge the people listed here, again, our VDRS programs out in the field as well as the vital stats, coroner/medical examiner, and law enforcement staff in the states as well as my CDC colleagues listed here. Thank you.
STEVEN SUMNER: Hello, everyone. My name is Steven Sumner and I serve as senior advisor for Data Science and Innovation at CDC’s Injury Center. Thank you so much for inviting me.
Today, I will be talking with you about youth suicide cluster investigations and in doing so, highlight the role of death data and other data sources as well as talk about challenges and opportunities with mortality data as it pertains to youth suicide cluster investigations.
I am going to do so by walking through a case study of a fairly recent youth suicide cluster investigation that CDC supported in Ohio. This example provides the opportunity to discuss in sequence the key considerations when using or applying death data in field investigations.
The specific suicide cluster I will be talking about began in August of 2017 in Stark County, Ohio. Stark County is located in Northeast Ohio and lies in a relatively rural area compared to the more populous cities of Cleveland, Pittsburg, and Columbus, which surround it.
And over an approximately eight-month period, Stark County experienced 12 suicides among middle school and high school students. Three of the suicides occurred in a span of less than two weeks and deaths were clustered in a relatively small number of school systems.
As typically occurs, following intense concern by the local community, local school officials, and local media, the Ohio Department of Health was engaged and formally requested CDC support for an epidemiologic investigation and recommendations.
In this Ohio community, as in most across the US, there was no formal cluster monitoring system in place. Nationally, there are perhaps a small number of well-resourced local health departments that do employ modeling approaches to try and detect youth suicide clusters. The exact number of health departments doing this is not exactly known. And cluster detection systems have largely historically not been implemented at the federal level simply because CDC does not have access to death data until many months or years later, meaning that real-time cluster detection systems simply cannot be built, given the current lag of mortality data.
Most youth suicide clusters in the US, if not all, are currently detected as a result of local communities, raising concerns about what they feel is an anomalous pattern happening in their community. It is not known how many suicide clusters may be missed in the United States.
One of the first key questions in an investigation is whether there is indeed a suicide cluster that is happening. In plain language, a suicide cluster would be defined as suicide deaths occurring in time and space more frequently than would be expected by chance. However, this is a complex question in and of itself to answer.
In Stark County, Ohio, the community had historically experienced about 3 youth suicide deaths per year so 12 deaths in an 8-month period were clearly alarming. In our investigation, we applied a rather sophisticated test, the Anderson-Darling Test, to demonstrate that the number of suicides were indeed statistically significantly occurring more frequently than would be expected from a random process.
In this relatively complex figure, you can see that the suicides that occurred during the cluster, which are colored here in blue, all fall to the left of the line, indicating that they are occurring at a rate much faster than would be expected by chance. The black dots in contrast are from suicide in Stark County prior to the cluster date. Some fall to the left of the line, some to the right, indicating a more balanced distribution around the expected rate.
This mortality data, however, was only available to CDC because it was shared directly by Stark County with CDC officials. As I mentioned, CDC Injury Center scientists typically do not have access to real-time death data to be able to do this sort of monitoring and analysis for communities.
An additional important point is whether in the absence of real-time death data whether clusters can be detected from other measures such as emergency department data. This figure shows emergency department data provided by Ohio to CDC for analysis. Stark County is outlined in black in the Northeast of the state. While Stark County experienced a growth in the percentage of youth emergency department visits related to suicide over time, it was by no means the county in Ohio that experienced the highest such percentage growth over time.
Suicide clusters do not always incur in places with the highest preexisting suicide-related morbidity or mortality and instead are a challenging phenomenon because of how they can spread in any community.
This figure shows data specific to Stark County with the red bars being deaths and the teal line being the percentage of youth ED visits related to suicide. We can see that ED visit trends do generally appear to correlate with death trends, which means that more work in this area of modeling of morbidity data might help detect emerging clusters early as well as monitor for their resolution.
The second major phase in an investigation after it has been determined that a cluster is indeed occurring is answering key questions for the community such as what caused the cluster and what are the ongoing risk factors that are perpetuating it. This is where death data presently has significantly limitations. Even in the case of Stark County, which experienced 12 youth suicide deaths, it is exceptionally difficult to study risk factors in a statistically significant way without count data that is much higher.
Furthermore, examining death data alone in this sort of context does not provide a comparison population to understand how the risk factors seen among decedents compare to the broader population of youth. The fundamental challenge is that death data while very comprehensive in certain regards such as autopsy examination or toxicology testing sometimes does not provide data on certain circumstance information of public health important.
In fact, as Janet presented on detailed circumstance information is only systematically compiled through the violent death reporting system processes conducted by state health departments several months after a death has occurred.
I want to make an additional important point about other challenges related to death data. In some investigations I participated in, particularly where the jurisdiction involved is experiencing a large number of deaths, processing of data by local offices can sometimes be delayed due to the fact that the sheer volume of work exceeds capacity. In these instances, CDC scientists have helped manually abstract information from incomplete records, trying to help bring the epidemiologic picture up to date in as rapid fashion as possible.
I will talk now about several new things we did in Ohio during this investigation. But most uniquely for this investigation, we did a large-scale digital data collection across the county to address some of the information challenges that we had.
What I mean by that is that in a very rapid fashion, we, along with the Ohio Department of Health, deployed a secure survey in schools across the county that asked students about cluster-related risk factors, screen for suicide risk, and paired this with options for linkage to care. This complemented our investigation of traditional data sources such as death data and helped fill in gaps for questions that were not answerable via these traditional sources.
Ultimately, over 15,000 students in over 30 schools were reached, which was our largest field response to date in the Injury Center. The reason for this large scope, as I will explain, is that schools across the county were highly connected in friendship networks. Deaths were spreading to other schools and school officials across these schools desired participation for their schools.
The digital data collection, which was administered in school on student chrome books, was a secure survey posted on the Ohio Department of Health REDcap server. Because 34 different schools were taking the survey, CDC prepared an analytic script, which analyzed the real-time data being received and prepared customed reports for each school system, highlighting school-specific risk and protective factors, measures of suicide ideation, and placed the results in context of other participating schools and Ohio, in general. Schools provided onsite counselors and mental health support for any students desiring additional help.
I will not get into all the results of the investigation because there are over 100 pages in the final report and two published studies that resulted. But what the real-time data collection allowed us to do was to assess in this context things such as several of the perceived traditional sources of contagion, things like exposure to suicide news or vigils as well as risk factors and other things not available through other means of data collection.
For example, what we found in this study is that exposure to suicide cluster related content on social media seeing material but really more so posting about it was one of the greatest risk factors. You can see here in these figures that students who had suicidal ideation – the red dots were more often seeing as well as posting about suicide content in a variety of social media platforms while the cluster was ongoing, the most common being Snapchat and Instagram among this particular population.
We were also able to explore in depth a wide variety of risk factors, which do not have as much data on them in these sorts of settings such as ACEs and substance use.
I alluded to this previously. But in this data collection, we were also able to ask students about their friendship networks across schools. I will not get into the full complexity of this component but each of the dots in this network figure represents a school and we were able to map out how schools were connected to each other through the student friendship networks.
Having a close friend who died by suicide is a significant risk factor and this was a first step and a new direction for us of incorporating more robust network studies into cluster investigations.
The final stage in a suicide cluster is essentially monitoring for resolution of the contagion. Since deaths are still a moderately rare event on a population level, many communities do not know when to feel relief in the sense that another death may happen is a very real concern for school systems, parents, and community members.
Furthermore, if another death occurs, the community knows about it before health department officials making the role of public health more reactionary than ideal. This is where potentially new data sources such as morbidity data such as the real-time emergency department visit data available to Ohio and CDC or other proxy measures for mental health offer value and may help public agencies monitor for general improvement in the community in a very real time and proactive fashion.
In our work in Stark County, we found that a large proportion of students expressed suicide-related thoughts. These percentages are often more elevated than many people unfamiliar with suicide-related studies expect. However, they highlight an important consideration for suicide cluster related information in that death data represents nearly the tip of the iceberg and many more individuals are in a population that are at risk. The size of the population I think should be measured and adequately scoped prevention strategies implemented.
There is some academic debate about whether screening for suicide risk should be done, particularly in settings of limited resources. But Stark County stepped up the level of school counseling and mental health support to match the number of students needing assistance.
In closing, I would like to thank you for the opportunity to present. I hope this presentation has helped to illuminate how death data is used in suicide cluster response as well as provide some context on the limitations of death data and the need for supplementation with other data sources or opportunities where we could potentially improve the use of death data in suicide cluster response.
KARIN MACK: Thanks, Abby, Janet, and Steve very much. This is such a great depth of experience in this area. Thank you for sharing all that. I would also like to thank the three of you for all you have brought into prevention efforts.
While people are thinking about questions, I am going to echo some of the comments from earlier to make sure that you take care of yourself. I know it has been a long day. All you wonderful people, start thinking about your questions but also think about taking care of yourself at the end of the day, getting outside, or in your room, acknowledging how you are feeling and focus on doing some centering and taking care of yourself.
The first question came from the chat, and I think it speaks to each of the presenters to think about suicide amongst rural youth. What does that data look like or for an investigation, what was that like in a rural area? Are there things you think need to be added to data systems? Did we capture everything and how do these systems tell us about rural suicides?
ABBY COLLIER: While we, at the national center, cannot share a state or a substate jurisdiction, we do elect (inaudible) suburban, rural, frontier community. Any of our data can be analyzed by location type. We are getting ready to release a brief on health equity considerations in suicide and geography is a big component of that brief.
STEVEN SUMNER: I guess what I can add to that is that first of all, in terms of suicide clusters, suicide clusters really can occur in any setting, urban, suburban, or rural. I think particularly in rural communities, as I mentioned, one of the first questions is, is a youth suicide cluster occurring? That can be particularly difficult to determine with death data because population counts are lower in rural areas. I think improving suicide cluster investigations by triangulating information and a variety of other data sources such as emergency department data or other measures or morbidity might give us a more timely way to really answer for that community is something happening that is out of the norm.
KARIN MACK: Janet, anything you wanted to add?
JANET BLAIR: In NVDRS, we can definitely look at this issue and some folks have looked at it. We can actually look at this issue in NVDRS. Some folks have looked at it in a study that was done years ago and found that folks in the rural areas, decedents in the rural areas did not have mental health care as much. It could be an access to resources issue. More suicides related to firearms. That analysis probably does need to be updated but we do have that ability to look at that in NVDRS given that we collect Census information.
JOHN BLOSNICH: I wondered if you all could speak to an issue that is different about how we collect data for this. It is proxy report or it is passive – several people kind of categorized it but it is different from how we think about public health service where we are actually asking a person to say yes or no to a question. Abby, it may be comes up in how you all gather data. At least in L.A. County when we do the review teams, we have to have a specific way of categorizing whether we have an actual denial of something. No, this risk factor was not present versus it is undetermined. But is it undetermined because we did not bring it up and discuss it or is it undetermined because we discussed and we just do not know? It becomes this question of while we are trying to find the things that we do know of being where of how we document the things that we do not have and do not know.
ABBY COLLIER: A common concern. I think we try to differentiate by saying you mark unknown if the circumstance around unknown is this versus leaving it blank if the circumstance is that. That relies on the user to read the data dictionary and they probably do not do that very often.
I think what we have started doing more frequently is using the data and clearly identifying those limitations and trying to dig into what are the barriers around unknown and really getting them to document them in a very concrete way. We struggle with the same thing I would imagine NVDRS does as well.
JANET BLAIR: That is great. That is a great segue. We struggle with that in NVDRS. Particularly, we are looking at mental health treatment, things like that. We know that it is probably an undercount just because coroners/medical examiners, not to generalize, but you all talked about the difficulties with getting medical record information. These investigations are moving. You want to get these cases closed out. It is likely underestimated. Do you have some categories that we use for unknown?
While we are talking about items that are missing or unknown, sexual orientation and gender identity variables have high levels of missing. Just for whatever reason that information is just not getting in the system. I know that the previous medical examiner speakers have talked about some of the challenges with that. If any of you actually at this workshop have ideas, I know Kathy presented to our NVDRS group. Dr. Pinneri presented to our NVDRS group about ascertainment of SOGI data and just what a challenge that is.
I think that there are things that we can think about. We can think about supplementing. The issue with supplementing NVDRS data, which you may have this issue too, Abby, is that you may be adding on to your data collection time then. Can you try to get medical records and then you have to go through another hoop with how do you do the MOUs for that or MOAs for that?
But one thing that I wanted to mention about NVDRS is that it is a tradeoff, waiting a while for the information. You get that rich circumstance information that you are unable to get from some other data sources. But I know we talked with our VDRS programs too about this and what are you seeing out in the field. When we make changes to the NVDRS instrument, we always confer with the states and ask them how often do you see this. Do you think that this would be something that we could get?
For example, using an example of people dying by suicide in cemeteries, it got to the point where so many abstractors were seeing that until now cemetery is its own location in NVDRS.
And then I did want to mention self-care. We encourage our abstractors to practice self-care. We have presentations on self-care in all of our site visits. Someone from our CDC employee assistance program gives a presentation. A lot of states have implemented self-care as part of their routine because they know that these narratives can be very disturbing through them. Sometimes people may – an abstractor may even see a record of someone that they know. Usually they are told do not abstract that case. I just wanted to mention that self-care is a priority for our VDRS recipients.
KARIN MACK: That is great information.
A follow-on question before I get to the people who raised their hand. There was another question in the chat about medical conditions and how they are captured both in NVDRS and in – Abby, in your data as well. Some medical conditions, sexual health, sexually transmitted infection. Are you able to capture that kind of information?
ABBY COLLIER: It is there. The ability to capture it is there. Whether or not the data are filled in I think is another story. We did add something to the case-reporting system I want to say three years ago that asks in months when the child’s last touch with a pediatrician or a primary health care provider was. It will be interesting to see what those data show. They can put it in there if they have access to it. Yes and no.
JANET BLAIR: With NVDRS, again, it depends. We do have the ability to capture mental health diagnoses and we provide standardized guidance. For example, if the toxicology comes back on a victim and that victim has anti-depressants in their system, they should not be – the depression should not be endorsed or checked in the NVDRS instrument just because anti-depressants can be diagnosed for different conditions, for example, such as anxiety and what not. We have very standardized guidance that we give about that.
Other conditions, things like HIV or chronic pain, different things like that, are going to show up in the narratives if they are considered to be a circumstance. If that is in the coroner/medical examiner or law enforcement narrative, that abstractor is going to put that information in the paragraph’s description. When the researcher does a scan of the narratives, those key words will show up. And again, that helps to inform us. If there is something that we are seeing a lot, then we may decide and may talk with the abstractors and decide maybe this needs to be its own variable in the system. That is how we collect information on diagnoses.
JOAN LUBY: This is a question for Steven because I just think the issue of suicide cluster (inaudible) is a really interesting one that might be in these escalating rates. I was wondering if anything is known about developmental variation of that trend? We tend to think about it as being salient in adolescence when you get into tight, social groups, but has anybody looked at it in youngers or olders, and also just the question of social networks generally, and how that drives it.
STEVEN SUMNER: I think the short answer is that it is a very important area where a lot more research is needed. As I mentioned, it is unclear exactly how many clusters occur in the United States. People generally put the contribution of clusters to suicide, that is in total to be less than 5 percent. But there are some countries in the world, for example, Australia that have done some more comprehensive evaluations of (inaudible) across the country to try and quantify the total number of deaths. I think that more work is needed in the US to really get at some of those detailed questions about – for example, what you just asked. What is the contribution of clusters to the rising rates of suicides that we are seeing among young people --
KARIN MACK: I will also say that we do get requests for technical assistance for clusters of all ages. We had requests about middle-age white men, for example (inaudible). You do not always go into a state to do an investigation. We do get requests for assistance.
KATHRYN PINNERI: I just wanted to say that our child mortality (inaudible) works very closely with the death investigators in my office to the point where we do not know what you all need, the same with NVDRS. We do not know what you all need until we see it. They actually provided packets. Our investigators fill out the packets before the meeting and provide them to the primary abstractors for them to already have that information so we can have more of a discussion about the child instead of just looking for data points. It has proven to be much more effective. It is a little more time consuming, but we just stick the packet in our pediatric folders and it is there and it has to be filled out by the investigator assigned to the case. It has really helped move our process forward.
The same with NVDRS data. When we know what the questions are to ask and we embed them within our case management systems, eventually someone is going to start asking them and clicking the boxes. It’s that communication that has to occur for us to give you what you want in a more timely and efficient manner. It makes it a little easier for us in real time to do it than a year from now when the abstractors are going through everything or like hey, child fatality review. It makes it more real time, more efficient, and it guides our investigations towards the data collection that will hopefully work towards prevention and intervention.
ABBY COLLIER: Just thank you for being such champions for this work. If I could take you and replicate you.
JOYCE DEJONG: I think I am somewhat echoing what Kathy is saying too. Some of these really – the questions about did they have a sexually transmitted disease, did they have this, did they have that, when we are looking for cause and manner, which is typically we are not always searching very deeply for – especially in the youth suicides, we look at it more because of the CDR process probably. But we are not – because if you look at the big picture and the incredible shortage of forensic pathologists and the staff shortages that we all face that you are just – those are great and invaluable information, but we are also restricted by not being able. I would love to just be able to say here are all the medical records. You guys figure it out but we cannot. We are not allowed to share that to give those documents over. Any kind of solution, the packets, even in the larger offices, embedding the abstractors in our office so that they are part of our office and then they would have access if there were mechanisms for that too that would also benefit.
I think oftentimes those data points are there. They could ask for the information if it is not there. But we are somewhat restricted, but I think largely motivated, especially in a meeting like this where you say this is all important. Anyway, that was all.
ABBY COLLIER: I think you hit on a really important point though that often in our surveillance systems, we want to try to collect everything and we are balancing competing needs with researchers, funders, and then trying to get stuff that our field can actually collect. We have an entire database that we call our wish list that get down to such a finite detail that I think fatality review teams would hate me if I tried to ask them to collect that. They just do not have that level of detail and it is hard to find that balance around what is available and what is not because there are so many competing needs with this kind of surveillance.
JAN GORNIAK: -- and Abby and Dr. deJong are saying. It is about managing expectations also. We know it is a suicide because of the scene. And then sometimes also that other information may not be known initially and that puts in an investigative report. And the doctors are going to be reading the medical reports and all that stuff and that does not get into the autopsy report or to the (inaudible) because it is not “important” because we are more into the what and the how and not the why. I think managing everybody’s expectations would help.
PARTICIPANT: I might take back my comments about replicating you –
KARIN MACK: Related to that, I had a question for people in the room about – start out about software and connectivity issues and sharing information. It sounds like there are other ways that you are going into a room and you are talking through a case. I am wondering what your most important partnership and other pandemic lessons that were learned in terms of data and sharing and capturing data. Have things changed in the past couple of years? Do you do things differently? Share data differently? Have things regressed, progressed?
JOYCE DEJONG: I think that have improved. I am the medical examiner for 12 counties. That means I have 12 CDRs, 12 child death review teams. And we also have elder death review teams. And we are starting suicide death review teams and opioid-related. I cannot go to all of these. But the virtual meeting has significantly improved our attendance at CDRs. I think that has been really valuable. And now and then they talk about going back in the room. Some of my counties are 200 miles away. I made a point to go to the CDRs. I would lose a whole day to go to a one-hour child death review because I thought – I tried doing it by phone, on speaker. I could not hear. I did not know what was going on. When they went to virtual, it was like we are all there every time. I think that that has improved it.
ANGIE HAYES: I manage the team in Harris County. When we went virtual, we went from reviewing 12 to 15 cases a meeting to reviewing 66 cases a meeting. We had great attendance but we did not get anywhere with it. That is an exaggeration.
PARTICIPANT: -- people talked too much.
ANGIE HAYES: In our meetings, you can have – person next to you and have that and it is a little bit easier to manage the larger conversation in person because somebody starts to get a little repetitive or talking over each other. It is easier to rein that in in person than on Zoom because you do not have that delayed voice of hey, stop, let us move on. It is easier to manage the people. In a smaller county, that might work because you have a couple of deaths to review. We have four-hour meetings monthly and we are behind.
We are reviewing 2020 right now because Harris County is enormous. I will actually introduce Harris County to you all tomorrow when I get up there. Some of these things like what Montgomery County does, we are not able to do in Harris County. We talked to our ME’s office about getting more details and that kind of stuff, and they were like it does not affect us. You all have fun with that. A month ago that was the call we had. It was not to be rude. Our office has 8 pathologists and fully staffed for our population would be 15. A huge portion of that is capacity.
KATHRYN PINNERI: If you shifted that burden to the investigators, which is what we do, you will have more success. And we went to the packets because of Zoom just like – give the packets to the investigator.
ANGIE HAYES: We cannot get through – we will talk.
MARGARET WARNER: Karin, you asked about did COVID change data sharing. I did want to make sure to give a shout out to the vital registrars and people (inaudible). We have a lot more provisional information out there in a much more timely way. WONDER, which of you may have seen – if you work with WONDER, we have the provisional data out. There is a six-month lag, I think, for the suicide deaths. It is really a huge improvement over where we were five or six years ago.
And then the other thing I wanted to mention is that the drug overdose crisis has also impacted getting the information out more quickly and particularly in offices of medical examiners and coroners. Some of them have dashboards available and suicides are – primarily, those are for drug overdose, but suicides are more – were becoming more commonly reported in a more timely way from medical examiner’s offices. I think it is a combination of COVID-19, but also the drug overdose crisis and the public’s understanding of the importance of that information as quickly as possible.
JOHN BLOSNICH: The methods we talked about, whether it be CDR, the cluster surveillance and investigation or NVDRS. Kind of slipping away from the US, how are we different from or similar to any other international efforts around this? Can we learn things from other systems? Can we teach things to other systems?
ABBY COLLIER: We participate in an international group of child death review programs, called something a little bit differently. The most robust programs that are in the UK and throughout Europe.
I think the biggest thing from my perspective that they do differently – there are two. One is data collection. They have more – their access to records appears to be broader. And the second is that they engage the family and that is definitely (inaudible) from my perspective in the child death review process. Now, I am not at all suggesting a family member comes to the review meeting. That is not what I am suggesting. But it would be great to incorporate their view in some capacity because they know a side of the story that we do not know and that might not have been captured or available during the death (inaudible).
JANET BLAIR: What I can say for the NVDRS piece is I do not know off hand of any other systems that are like this, not to say that there are not but we have had inquiries from people in Canada and Australia about how did the system get implemented and how did you start up and what is the approach. We have provided some technical assistance to public health people in those countries.
KARIN MACK: -- two virtual questions with Jane.
JANE PEARSON: Thanks. I actually have two questions for Steven. Since you have done some research with Bark on the social – the software monitoring in schools and you were talking about the secure servers with Chromebooks. Have you considered what is popping up on those monitoring systems to look at contagion in schools? That is one question.
STEVEN SUMNER: I guess I will take that one first. I think your question is generally around new data sources via new products and systems that schools are implementing. Can those be important for understanding suicide? I think the answer is generally yes.
I think there are – no particular product has complete and universal coverage. For example, in the case of the study that Jane is referring to, we collaborated with the school safety monitoring software program to try and understand what were the risk factors that were happening in a very objective way in student’s lives as measured through this objective data that school systems have and school safety monitoring software collects about what are things that students are exposed on their computers whether that is bullying, sharing of suicide-related content, sharing of violence or drug-related content. We were able to do a very detailed and I think useful case control study to understand the risk factors better with a very granular time scale. I agree with you, Jane. I think it is an area where – keeping an eye on new data sources that are popping up is a really important thing.
JANE PEARSON: And then also seeing what the school does with the data like do those kids get referred. You were looking at the front end like what was contributing to the kids’ reports of suicide markers or whatever is on the software and then what did the school do about it afterwards. I will stop with that question and let other people talk. Thanks.
KARIN MACK: Other online? Stacia?
STACIA FRIEDMAN-HILL: I am a program officer at NIMH in the Division of Translational Research. My question is – first, I wanted to comment that I have gained a new appreciation for all of the work that goes into creating these records, investigating, collecting the data. It is very much an intensive process that I think I did not have an appreciation for before.
My question is about once the data is collected and put into the various reporting systems, it is a more of a data science question for researchers who would like to use this information in the research. What kinds of interoperability is there between the different systems? Are there ways of cross-referencing the same child, for example, in multiple systems or using geocoding and looking in order to understand their locality better, how much violence there might be in their community, what particular public policies might have been in their community? Are there efforts to make that easier for researchers to get that interoperable picture?
ABBY COLLIER: I have an easy answer, which is it is not possible in our data at a national level. The state or substate jurisdiction can do any of that because they have access to all of those identifiers. I will share that Cook County, Illinois is getting ready to really release some really exciting dashboards around infant sleep that we hope catalyze folks into thinking about how they can use this data at the state and local level. But there is a lack of resources. It is expensive to do the kind of research that you are suggesting.
If anyone is ever interested in that in a specific state or you are open to a state recommendation, you can always reach out to us and we will connect you with that state. But we cannot do it. We cannot grant you identified access.
JANET BLAIR: I would just echo what Abby said as well. Within NVDRS, we cannot do one-to-one linkages either because we do not have personally identifiable information. The state has that information. They do not enter it into the system. But the VDRS programs at that level have that information and they have done some really interesting linkages with juvenile justice and other groups. That is where they can get the most out of the matching because they can do that one-to-one match.
KARIN MACK: Deb, I do not know if you want to mention anything about any of the linkages with veterans or with the CSP programs –
DEBORAH STONE: CDC - a few projects where we are linking. There is one project with the Department of Defense where we have been able to link data from the DoD suicide event report called the DoDSER to NVDRS data. That is through a probability linkage. There is another project out in Washington State where folks are looking at data linkage for across social service data, justice data, emergency department data. They have a whole range of different data sources that they are looked to mortality data to morbidity data across these different systems.
As far as our comprehensive suicide prevention program, that data is mostly states who are part of that are looking at using syndromic surveillance data but it is not data linkage. But it is providing a really helpful, near real-time data on suicide morbidity, suicide attempts, suicide ideation, and things like that from emergency department data across a range of populations. There are some really nice pilot projects and other linkage projects going on that I think that will inform prevention.
KARIN MACK: Maybe we have time for one more. I hear that people need more staff. You need more people to do the work. Other than that, what other resources do people need – some of the discussion for the end of day tomorrow. What types of things – is it training? Is it better information, other specific resources? Is it really just a better research foundation? What sort of things would help you in doing your job better? I liked Abby’s concept of useful and nimble data. If that is what we are trying to get to help with prevention, what sort of things do you need? What do we wish were better in your states or communities? What would be better at the federal level? What could we change? Any thoughts on this – eight hours into a long day.
KATHRYN PINNERI: I have a comment real quick. I think that the delay in data gathering is significant. I think a six-month lag from when you request the data so request in June the data from previous six months as opposed to waiting to the end of the year to get all of the data for that year. I do not think we are at a point electronically where that is necessary. It used to be when it was paper and you had to go and gather all this stuff. But I think you can request the data throughout the year. The formats are the same but just obviously keeping in mind that probably six months is going to be the best thing to make sure that you get cases completed because toxicology, we are all busy. We are all short staffed, all these things that we have talked about. But I do not think you have to wait for a full year’s worth of data to start your data abstracting.
JANET BLAIR: I just wanted to mention to that note, I wanted to talk about some of our data modernization that we are doing with NVDRS. We do have software that we use where the abstractor logs onto the program and the web and puts the data in. It is an encrypted system. We are developing a rapid reporting feature. What that means is that the VDRS program can go in – for example, the VDRS program wants to take a look at the deaths that occurred in 2022 before close out. They want to take a look at all of the suicides for that year 2022. We have not had our close out yet.
With this rapid reporting feature, they would be able to print out select characteristics and mini reports so that they could provide that information for their state so they could give it to CME partners, LE partners. They could see what is going on before the actual close out. It comes with caveats because you may not have all of those pieces. You may not have the CME report or all the toxicology back. You may not have the LE report. It comes with caveats. But it is going to make it easier for states to be able to report out earlier with all the caveats so you do not have to wait all that time for close out. We are working on it.
And then the other issue is one of interoperability. How do these systems all play well together? We have an import feature where the abstractor can import information from the death certificate into NVDRS. I know that CMEs have different systems that you all use, different electronic systems. Law enforcement has different electronic systems. How do we get all of those systems to be harmonized and maybe standardized systems? There may be different vendors, but standardized systems that can play well with our surveillance system. That is a challenge.
ABBY COLLIER: Let me just echo. I think data modernization is a huge need just across child death review staffing such a barrier. And then also, if you want consistent data, we have 2300 users, give or take a few. There are some differences in how people answer the questions. If we were getting it reliably from the same source it would cut down on the noise we see in our data.
And one thing we did not hit on is that if we were able to do some of these data linkages, we’d be exposing our teams and our staff to less of this information. If the abstractor does not have to read 70 pages of our report, the impact on them over time could be significant and I do think there is a workforce component that we do not think about as often around data modernization.
KARIN MACK: I am going go pass it over to Brendan who gives a final --
BRENDAN WEINTRAUB: Hi everybody on the call. We have one last Mentimeter today. I am sharing that now. It is going to be the same as before in terms of our Menti. What impressions are you leaving with today? If you can just go to www.menti.com, code 31733. We see great minds in the room, hopeful, better knowledge about warning signs for suicide, it takes a village. It is optimism. It is great to hear. Please continue putting these up on Menti. This is going to be open and it is really going to help inform some of our discussions as well tomorrow.
I am going to stop sharing and pass this over to Lisa to wrap up for the day.
Agenda Item: End of Day 1 and Preview of Day 2
LISA COLPE: Here we are at the end of Day 1. Thank you so much for being here with us today and virtually. First time holding this kind of meeting in this building. We had a few hiccups this morning and we are working on it. I think overall it was a tremendous success.
We heard about the death investigation process and its challenges, some information from the child research community and their findings and how additional data is collected by via child death review committees and the NVDRS. We also had some discussions about how can we help you meet your challenges. What kinds of things could we do differently?
Tomorrow, we have another great day planned. We will be talking about some special risk populations and have some experts in those risk populations that will be telling us a little bit about their perspectives about suicide in those populations.
And then at the end of the day, we are going to have some great talks about what states are doing with all of their data. Everybody does things a little bit differently. But we will be hearing about the variety of ways that states pull their data together. And then finally at the end of the day, we will do a little bit of a summary and then we have a lot of our federal partners here wanting to tell you about the resources they have or things that are coming up, research opportunities that are available. It is going to be another big day tomorrow.
I will tell you that since this is the first time we were doing this, I built in a little time at the end just in case we started to slide. We are going to be able to end early today and welcome you all back tomorrow morning at 9 a.m. to start all over. Thank you again.