One of the key treatment interventions is determining the people with the highest risk potential for the most intensive treatment interventions. The treatment outcomes in terms of averting aggressive and suicidal behaviors are generally good. Given the relatively rare occurrence of aggression or suicide post discharge the actual power of the treatment intervention is unknown. The potential severity of outcomes precludes any placebo controlled clinical trials. No human subjects committee would authorize a placebo arm and since many patients are on involuntary status or court holds. No probate court judge would go along with it either.
The March 2014 edition of the American Journal of Psychiatry has some the most most extraordinary content I have ever noticed in that publication. Among the articles is a paper called "Association of Violence With Emergence of Persecutory Delusions in Untreated Schizophrenia". It adds significantly to the literature on psychosis and violence. The study focuses on the United Kingdom Prisoner Cohort Study and it looked at risk factors for future violence in prisoners who were incarcerated for a violent crime after they were released. It is a study that could be done on patients who were acutely hospitalized and released because of the naturalistic design and use of nonviolent participants as a comparison group. That authors were interested in looking at whether the presence of psychosis predicted future violence and if there was any specific pattern of symptoms. They were also interested in looking at the issue of whether or not treatment was helpful.
The sample consisted of 1,717 prisoner screened at baseline and 967 followed up (787 men and 180 women). Selection was based on incarceration for at least 2 years for a violent crime and release date within 12 months of the start of the study. All participants were given a number of structured research assessments to establish diagnosis. At follow up, the diagnoses of the patients in the study included 94 meeting diagnostic criteria for schizophrenia, 102 for drug induced psychosis, and 29 for delusional disorder. Only the subgroup with schizophrenia scored higher on psychopathy scores. Violent behavior at follow up was established by self-report and a national computer police database that classified violence against persons. According to that database 22.9% of participants were violent between release and follow up (mean 39.2 weeks).
In terms of the relevant results, the delusional disorder and drug induced psychosis subgroups were no more likely than the the participants without psychosis to be violent at follow up. Persons with untreated schizophrenia were more than three times as likely to be violent that the non-psychotic participants at follow up. In that group those with persecutory delusions were more likely to be violent than those with other symptoms of psychosis. The authors briefly review the indirect evidence supporting their findings including treatment non-adherence and risk of violence, risk of violence at first presentation of treatment rather than subsequent episodes, and psychosis as a risk factor for violence. They point out that to their knowledge this is the only study of violent recidivism in prisoners that looks at the issue of psychosis as a risk factor.
The actual treatment provided in this case was critical. In terms of violence prevention any treatment provided in prison only or in prison and on release was effective in preventing violence. They point out that identification of more people needing treatment by their study methodology may have led to more active treatment of study participants. They quote data on that fact that in prisons in the UK only about 1/4 of prisoners with severe mental illnesses are identified by mental health teams with that specific function and that of those identified only 13% are accepted into case management. Overall in the UK less than 1/4 of prisoners who screen positive for psychosis are given a mental health appointment at the time of discharge.
The accompanying editorial by Large is interesting in reviewing the issue of screening versus not screening populations for psychosis and whether that prevent violence. Several studies have concluded that "risk assessment is insufficiently sensitive to provide a basis for protection of the public." Without looking at all of the references (I would expect to find significant flaws) the issue is really not a screening issue. This study happens to appear like it is a screening, but the diagnostic approach is probably much more vigorous than most assessments in correctional settings. The issue is that you have a person sitting in front of you telling you that they have persecutory delusions and are at risk for continued violence secondary to those delusions. There is also a significant subgroup who are at personal risk for self harm related to these delusions that the authors either did not find or they did not comment on. The Large commentary also focuses on antipsychotic medication as the treatment for psychosis and in the UK psychotherapy is also a treatment modality. He makes the observation that treatment across the entire spectrum is important in that less treatment in the currently treat group will also result in more violence.
This study is useful in the US for several reasons. County jails have become the largest psychiatric hospitals in the United States largely as a result of government and business policy. Inpatient units may be useful for acute violence but there is an uneasy relationship with county jails. Hospital policy may result in suicidal and acutely aggressive psychotic patient being treated in jail settings and using methods that would be seen as completely inappropriate in a medical or psychiatric setting. Psychiatric follow up in jail settings is often fragmented and it is not uncommon to see medical treatment started and stopped based on the availability of medical staff or prescription medications. I would consider the UK to be much more enlightened with regard to mental health policy than the US and to have more medically based resources for anyone with a psychosis diagnosis. I can't imagine follow up numbers from American jails being any better than they are in the UK.
All of this creates a problem for the person with psychosis, persecutory delusions, and violent behavior. The focus of much of the literature seems to be protecting the public from them but when you are their treating psychiatrist the arguments you are making to them is to protect them from their delusional thoughts. That will not happen in a rationed, carved out environment that has shifted progressively more care for the severely mentally ill to correctional settings. The other interesting cultural phenomenon is that there is no coverage of this study or similar studies in the press. Their bias seems to be to look at the sensational results of psychosis associated violent crime, suggest that more treatment might be needed, attribute causation to being in the wrong place at the wrong time, and suggest that we all need to move on (lurch forward?) toward the next catastrophe.
This study provides a platform for a better approach to public policy and a more patient centric approach to violence prevention.
George Dawson, MD, DFAPA
1: Keers R, Ullrich S, Destavola BL, Coid JW. Association of violence with emergence of persecutory delusions in untreated schizophrenia. Am J Psychiatry. 2014 Mar 1;171(3):332-9. doi: 10.1176/appi.ajp.2013.13010134. PubMed PMID: 24220644.
2: Large MM. Treatment of psychosis and risk assessment for violence. Am J Psychiatry. 2014 Mar 1;171(3):256-8. doi: 10.1176/appi.ajp.2013.13111479. PubMed PMID: 24585326.