How Not to Screen for Suicidal Thoughts: Lessons from Mountain View Police
A couple of days ago, Mountain View, CA police released the video footage of their interaction with Nasim Aghdam about 11 hours before she shot people at You Tube headquarters. It is a remarkably unremarkable video in the sense that the police are polite with her and she is polite in responses, and the interaction is uneventful.
Nevertheless, as a psychiatrist who used to enjoy teaching psychiatric interviewing to medical students and residents (and to other psychiatrists when helping them prepare for their oral boards), I was struck by how the police asked the questions to screen for suicidal or homicidal intent.
At 3:13 in the video, an officer asks, “You don’t want want to hurt yourself, do you? You don’t want to hurt anybody else? You don’t want to commit suicide or anything like that?” The 3 questions and their answers take a mere 7 seconds.
This way of asking questions about difficult topics leads the subject of the interview towards a particular answer, usually one that relieves the interviewer. If the subject were to have the slightest willingness to talk about thoughts of hurting herself or others, going through 3 yes/no questions in 7 seconds makes it clear that no one is truly interested in listening. I could have told them, by the way the questions were phrased, that the answers would be no.
I am not blaming the police here. The role of the police is different than the role of clinicians and even clinicians make these mistakes, often habitually. I have seen trained mental health professionals – psychiatrists, psychologists, social workers, etc. – do similar things and ask yes/no questions regarding topics that require more nuance and gentle prodding. I have seen them ask things like:
“Have you ever used illicit drugs, have you?”
“You have never been abused, have you?”
Asking these same questions in a different manner does not guarantee that the subject of the interview will be truthful, but increases the odds.
So how should someone approach such questions? The first step, of course, is to have an unhurried attitude indicating that one has time to listen. The next step is to get rid of the habit of ending such questions with “do you?” or “have you?” When the question is phrased this way, it’s hard for an interview subject to avoid the feeling that she will be judged negatively if her answer is in the affirmative. Third, it helps to phrase the question in way that doesn’t lend itself to a prompt yes/no response, i.e. don’t phrase as if it came from a checklist. It places too much responsibility on one question. Spread the assessment out. Finally, take the non-judgmental attitude one step further and attempt to normalize potential affirmative responses.
For example, instead of asking “You don’t want to hurt yourself, do you?”, it would have been impressive if the officer had said, “People disconnect from their family for a variety of reasons. Sometimes its because the family has hurt them. Sometimes it’s because they have thoughts of hurting their family and they don’t want to do it. Which one do you think applies to you?” This way of asking normalizes that in some situations a person might think of hurting others and does so in a manner that it doesn’t lend itself to an easy yes/no response.
A person may respond, “I don’t have such thoughts, I just wanted to get away from them.” I would follow up, “Bear with me here, but I have also seen some instances, not all, when a person is feeling so alone, that they have thoughts of ending their own life they think of ending their own life,” and just end without a question, letting the person respond to this statement according to how they are feeling at the moment. If she wouldn’t say anything, I would follow up and say,”When was the last time you felt so down that you came close to thinking that way?”
The point of this post is not to teach clinical interviewing to police, but to point out that there are ways of asking questions that increase the odds of a person opening up and there are ways that close the discussion too quickly. We must also be humble enough to know that that’s all it does – improve odds of, but doesn’t guarantee a more candid conversation. If it guaranteed it, mental health professionals would be fantastic at predicting violence by our patients.
Apollo's Minion 
Comments