Saturday, December 4

Coercion hardly leads to fewer murders

A fixed hospital bed with belts should prevent the patient from hurting himself and others.

We cannot lock in all patients who have a diagnosis of psychosis to prevent the five to ten murders that patients with psychosis commit annually.

This is a chronicle. Opinions in the text are the responsibility of the writer.

It seems to have become a collective notion that if someone has a mental illness and harms himself or others, the mental health service should have prevented it.

Society expects this part of the health service to deal with treatment, care functions, and social control. It is the latter function that makes people more ambivalent about mental health care.

As part of social control, mental health care also has an impossible task: predicting when people may commit irrational, dramatic and dangerous acts.

It seems to be a common belief that such actions could be avoided if everyone had the right to treatment and if those who did not want the treatment were forced to do so.

Look for explanations

It is understandable to seek explanations when there is a murder in public space.

In the wake of the tragic killings in Kongsberg, several psychiatrists and police officers have pointed to the adjustment of the Mental Health Care Act in 2017 as a possible explanation.

The change led to a higher threshold for involuntary hospitalization. This, in turn, made it more difficult to prevent patients from committing acts of violence, it is claimed.

It is not so easy to see the logic in this reasoning. We know that the number of involuntary admissions increased from 2016 to 2020, it did not decrease.

However, more importantly, the underlying cause is another. That is, it is almost impossible to predict (predict) an event with a very low incidence, such as homicide.

It has little to say if we adjust the law on mental health care one way or another, as long as we fail to predict who with a mental illness will commit murder.

We cannot lock in all patients who have a diagnosis of psychosis to prevent the five to ten murders that patients with psychosis commit annually.

Almost impossible to distinguish

Rather than starting with assumptions and accounts of individual observations, we would prefer to base ourselves on facts.

The death toll in Norway is very low. Every year, about 30 people die. This corresponds to a homicide rate (number of deaths per 100,000) of around 0.5. The homicide rate is surprisingly stable and not increasing (UNODC homicide rate data).

So far this year, there have been 22 homicide victims. In Sweden, the homicide rate is about 1.1. In the United States it is around 5. In some Latin American countries it is more than 100.

The international literature shows that between 5 and 20 percent of all murderers have a diagnosis of schizophrenia. In Norway, around 20 percent of all murders are committed by people who are known to be insane, most of them with a diagnosis of schizophrenia.

At any given time, there are around 16,000 people with a diagnosis of schizophrenia in Norway, between 25,000 and 30,000 with a diagnosis of psychosis. It is almost impossible to distinguish between them the five or six people who commit murders.

To avoid a homicide, calculations have shown that when using a good quality hypothetical prediction tool, 2,500 people with schizophrenia should not only be classified as high risk, but should receive treatment commensurate with being in this category (Large and employed, 2011).

Find what contributes to increased risk

Existing instruments have marginal predictive value with respect to murder / serious violence.

We wonder if it would be possible to improve the existing forecasting tools. In this context, we conducted a survey in 2016-2019.

If we could find out more about the psychological factors that contribute to an increased risk of murder or other serious violence among people with schizophrenia, it would be possible to make the risk assessments a bit more precise.

If improved predictive tools can help identify a single potential killer and save a life, then we must try.

Compared patients

The study consisted of comparing 26 patients with schizophrenia who had committed murder, with 26 patients with schizophrenia who had never committed violence.

Our starting point was that if these two groups differed somewhat from what we measured, then this could be used to improve our ability to predict who may commit violence against others. We measure, among other things, cognition, social cognition, and some personality traits.

We found that participants who had committed murder / attempted murder had great difficulties with cognition and social cognition. That is, how one perceives the world around him and processes the information.

They had significant difficulties with learning, memory, and problem solving, skills that affect our daily functioning.

May be a partial explanation

However, we found the largest group difference for social cognition.

Participants with a history of murder / attempted murder showed very significant problems understanding other people’s thoughts, intentions, and feelings. Therefore, they had a social disability characterized by a lack of perception or misinterpretation of social information.

A basically neutral action can be interpreted as hostile in people with such difficulties. Therefore, reduced social cognition may be a partial explanation for serious violence / murder.

We also received confirmation from previous research results that psychopathic personality traits are strongly associated with murder (Engelstad et al., 2018, 2019a, 2019b).


We knew from previous research that the strongest risk factor for homicide among schizophrenia patients is substance abuse. More than 70 percent of all killers with psychosis have a substance abuse problem.

Of the personality factors, it is psychopathy, impulsivity, and lack of perception that most clearly link schizophrenia with severe violence.

Furthermore, it is also worth noting that clearly most murders are committed during the first episode of a psychosis, before the person has entered treatment (Around 2018).

Most are never a danger

We hope that some homicides by patients with schizophrenia can be prevented if all factors that pose an increased risk of violence are included in the violence risk assessments.

The cognitive and social cognitive factors that we identified in our study should be included in these evaluations. Plus, research shows how important it is to get there early with treatment.

Since the myth that many people with mental illness are “time bombs” controls media coverage, we remind you that the vast majority of people with a psychotic disorder will never pose a danger to their peers.

On the other hand, they themselves are exposed to violence to a greater extent than others.

Leave a Reply

Your email address will not be published. Required fields are marked *