Warning: This post is about suicide and the assessment of suicide. It is intended for mental health professionals and people who not distressed by this topic. Avoid reading this if you find the topic of suicide distressing.
Suicide assessments constitute a major part of psychiatric
practice. According to standard
guidelines it is a recommended part of any initial assessment. Acute care psychiatry selects for these
assessments largely because hospitalization and crisis care is focused on it
and aggressive behavior. Over the past
30 years insurance companies and governments have made it virtually impossible
to treat people in a secure environment unless there is a risk of suicide or
aggression.
Most suicide assessments are taught as an exercise in risk
factor analysis. Patient traits, demographics, and diagnoses correlated with
suicide or suicidal ideation are collected across studies and applied to
current evaluations. Decisions about
treatment are made on that basis. The
decisions may also have implications about continued risk despite what is said
in the interview. It can be a basis for court
ordered involuntary treatment.
For example, let’s say I am asked to see an 80-year-old man
on a surgical service. He is there
because he tried to cut his throat and underwent surgical repair of his
esophagus and trachea as a result. He
describes feeling better at the time of the interview but says he has been
depressed for years. He lives alone
after his wife of 45 years died last year. At some point he noticed that there was a foul
smell covering his body. He thinks the
smell comes out of his mouth at night and covers his entire body. He is a heavy drinker and consumes 500 ml of
vodka per day. He prefers to return home
without treatment as soon as the surgery team clears him for discharge. His labs show elevated transaminases and prolonged
coagulation parameters.
This is an example of a person at high risk for ongoing
suicide attempts based on risk factors.
In this case depression, psychosis, alcohol use, a serious suicide
attempt requiring surgical repair, age, and lack of social support all define
him as high risk. It is unlikely that
any psychiatry service would discharge him untreated to go back home and
potentially experience the same series of events that led to the attempt.
That was state of the art assessment back in 1982 when I
started my residency and it is not much different now. A few months ago, I sat through a very long presentation
on an artificial intelligence (AI) based approach to suicide assessment. It consisted of analyzing the patient’s word
frequency during the assessment and deciding suicide risk based on that. The qualifier was that it was not a substitute
for clinical judgment. It reminded me a
lot of the quantitative electroencephalogram (QEEG) research I started doing in
1986. The technology claimed to be able
to separate psychiatric diagnoses based on fast Fourier transformation (FFT)
analyses of EEG frequency bands. The problem was the analysis also depended on clinical
features that had to be added to the diagnostic algorithm. It was not a true test without that
additional input. The AI analysis of suicide was no different.
The problems with assessments for suicide potential are
essential two-fold. First, the conscious
state of the individual changes and they go from a person who would never
consider suicide to one that would. Before
that change you are talking with and gathering data from a person who is not contemplating
suicide. Second, suicide attempts are
generally impulsive. Many people
interviewed after surviving a suicide attempt are glad they survived. In many cases they regretted to committing to
suicidal behavior almost immediately. A
good example are the young men who survive jumping from the Golden Gate
Bridge. They experienced instant regret
after jumping away from the railing. An
additional complicating factor is that the person sitting in front of you may
want to be released to make another suicide attempt and they either do not want
to discuss it or they want to conceal that fact from you.
I had all these things on my mind when I was doing consults
on medical-surgical patients at the hospital where I trained back in 1982. I was a first-year resident and my job on
this rotation was to show up and do all the preliminary evaluations on the consult
requests that day and then present and discuss them with my attending Dr.
H. Dr. H had been an attending for about
6 years at that point. She had returned
to work in the county hospital from private practice. I had worked with her for a few days and
things seemed to be going well. We
generally agreed on diagnoses and treatment plans and there were no personality
conflicts. That is about as ideal as it
gets for a resident. Then one day – Dr.
H showed me an interview technique that I never forgot.
I had just presented the case of a young man who had
overdosed on antidepressant medications.
He seemed mildly depressed and irritated. I ran down his history and probable diagnosis
to Dr. H and we walked in his room so that she could interview him. It went something like this:
Dr. H: “Hi I am Dr. H
and I am the staff psychiatrist here. Dr.
Dawson was just telling me a few things about what happened. Would it be OK if I asked you some questions?”
Pt: “Sure.”
Dr. H: (after clarifying
the demographic and medical data): “Can I ask you about the overdose”
Pt: “Sure”
Dr. H: “Do you
remember the details? Do you know the
pills you were taking?”
Pt: “Yes they were
amoxapine.”
Dr. H: “How did you
take them? Did you take them all at once or one at a time?”
Pt: “I was taking
handfuls. I would take a handful at a
time and rinse them down with water. It
was hard to do because they are large capsules….I had to take more and more
water and eventually stopped.”
Dr. H: “And what
exactly were you thinking at the time?”
Pt: “I was thinking I
wanted to die. I was thinking that I was
a loser and I wanted to die. I could not
see any future. I did some research on
this and knew that this stuff was fairly toxic and that if I took enough of it –
it would kill me. I was throwing them
down as fast as I could.”
Dr. H: “What happened
next?”
Pt: “At some point I started
to feel sick and I got really drowsy and passed out. The next thing I was
waking up in the Emergency Department downstairs. They had a tube down my throat and they were
giving me charcoal.”
Dr. H: “Looking back on what happened yesterday –
what do you think?’
Pt: “I would not do
the same thing again but it would not have bothered me if I succeeded
yesterday.”
Dr. H: “Do you feel
like a different person today?’ ….
The above exchange is a brief excerpt of the interview, but
it was not like my interview. I spent
about an hour interviewing the patient about depression, anxiety, and suicidal
ideation like they were all third person observable objective facts. He was
clearly less engaged with me than he was with Dr. H. When you interview someone from the perspective
of third person objective facts – you invite them to see the world the same
way. They become passive observers to
what happened to them. You can’t really
get to the change in conscious state or impulsivity that make suicidal states
unique. Dr. H went on another 20 minutes
getting every detail of this patient’s subjective experience of the incident. It was amazing and we discussed it when she
was done.
From that point on my suicide assessments were all based on
that phenomenological approach whether I was talking with people who survived
attempts or were talking with me because they feared losing control. I needed to know their emotional state and
what they were thinking. Even in those
descriptions there were conscious fantasies and defenses: “I was pointing the gun at myself but I never
pulled the trigger. It just went off in
my hand.” In the process I heard
hundreds if not thousands of reasons why people attempt suicide and exactly
what they were feeling and thinking at the time. In the larger scope Dr. H helped me focus on
the subjective. That is something that
you lose in medical school where there is an implicit emphasis on the objective
and subjectivity seems like a bad thing.
The reality is that subjectivity dwells within every classification
system.
My memories of the past are so vivid that at times I forget
I am an old man. I recalled the above exchange
with Dr. H when I was discussing phenomenologically based approaches to suicide
assessments with a new generation of residents. That happened just last week. I decided to look her up and see what she was
currently doing. I wanted to thank her
for the direction she gave my development and career. I found out that she died 6 years ago. Her obituary said she did not want a funeral. The family requested memories and
stories. I hope this blog serves that
function. She taught me about phenomenological suicide assessments when they
are scarcely written about to this day.
I am sure she taught many more people than me.
Passing an important technique along that you can’t find in a
book or a paper and making that accessible to a young resident who thinks he is
getting the job done is a great legacy.
Thank you Dr. H!.
George Dawson, MD, DFAPA
References:
1: Nelson K. ‘All I wanted to do was live’: After years of
debate, a suicide safety net for the Golden Gate Bridge is nearing completion.
Survivors say it’ll give many a 2nd chance at life. CNN.
November 19, 2023 https://www.cnn.com/2023/11/19/us/golden-gate-bridge-suicide-safety-net
