Murder and the mentally ill

By GEOFF CUMMING

Everyone knows society is more dangerous these days: there's violence on the streets and it's beamed into our living rooms. And it's no coincidence that the closure of Carrington, Tokanui, Lake Alice and other mental institutions led to this surge in murders, right?

Psychiatrist Sandy Simpson tests the theory on his masters students and the community groups he addresses: do they think the proportion of murders committed by the mentally ill has risen or fallen in the past 30 years?

"They all think it's getting worse."

But a study released today quashes myths about the risk posed by the mentally ill. It finds no basis to claims that care of psychiatric patients in the community has contributed to the increase in murders. It even suggests that the likelihood of violence by people with mental illnesses has fallen since "deinstitutionalisation".

"Community care hasn't failed," says Jim Burdett, who represents people in contact with mental health services. "People aren't being killed by lunatics in huge numbers."

The study, Myth and reality: the relationship between mental illness and homicide in New Zealand, examined the files of 1501 murderers between 1970 and 2000. It found 133, or 8.9 per cent, were judged not fit to stand trial or not guilty by reason of insanity, in line with rates for low homicide countries internationally.

But an interesting trend emerged. Although the number of murders has tripled, from fewer than 20 a year in the early seventies to between 50 and 70 in recent years, the number of killings by the mentally ill has remained steady - and is usually less than six a year.

The result: murders by the mentally ill have fallen as a proportion of total homicides from nearly 20 per cent in the early seventies to around 6 per cent today.

The study is one of only four in the world to look at time trends in homicides, and is touted to be the most comprehensive. Researchers satisfied ethical and privacy concerns to cross-match data from the police, coroners' office, Ministry of Health, Justice Department, Parole Board and New Zealand Herald archives.

Simpson, who led the research team, says it is significant that the study period coincides not only with deinstitutionalisation but an era of increasing social fragmentation.

Rising unemployment, poverty, marriage breakdowns, drug and alcohol abuse and increased exposure to violence are factors which could lead us to expect an increase in mental illness. But the 18 killers judged to be mentally ill in the past five years of the study period compares favourably with the 20 identified in the first five years.

"It almost makes me proud that, in a less hospitable world, people with mental illness haven't succumbed to these pressures in the same way that people without mental illnesses have by murdering more often," says Burdett, who himself has a mental illness.

The study for the Mental Health Research and Development Strategy addresses another myth: that community care raises the risk of being killed by a stranger who knocks on the door or "loses it" in the street.

Of 84 murders by strangers between 1988 and 2000, only two offenders were mentally ill. The flipsides: mentally "normal" perpetrators are more likely to kill strangers and the mentally ill are more likely to target loved ones and family.

Tony O'Brien, senior lecturer in mental health nursing at Auckland University, says the research confirms there's more reason to fear a drink or drug-affected single young male than someone with a mental illness.

"The fact that homicides by the mentally ill have stayed constant over this long period of deinstitutionalisation suggests mental health services are doing a pretty good job of managing the risk."

In the cauldron of emotion, science and politics that is mental health, there is reluctance to make too much of the findings, a mere stepping stone in research to improve the delivery of mental health services. The researchers admit the data has limitations and expect robust scrutiny from both the pro-institution lobby and mental health service consumers.

"I don't think you can say it gives community care a free bill of health," says Simpson, clinical director of the Mason Clinic in Auckland. "We're not saying dreadful things haven't occurred because of deinstitutionalisation.

"Nor are we saying there's not a case for more acute beds and some long-term beds."

When he started at Porirua Hospital in the mid-1980s, there were 1200 beds. Now there are no more than 800 nationwide. But Simpson says it's important not to romanticise the past.

"We would have two or three nurses in a villa with 35 people in it and one doctor in charge. The quality of care we could provide to people in terms of frequency of review and planning of rehabilitation was very limited indeed.

"Then when we shifted into the community the resources were never adequate to provide thorough care."

The failings of community care fuelled public suspicion that deinstitutionalisation was aimed more at cost-cutting than better service. Many services failed to come to grips with law changes which protected privacy and allowed patients not subject to compulsory treatment orders to refuse medication.

After the Raurimu massacre, and again after Malcolm Beggs was killed by his flatmate, coroners linked community care to the rising number of homicides and suicides. Politicians took up the cause, including former Prime Minister and psychiatric social worker Mike Moore, in 1997: "The care of acutely disturbed psychiatric patients outside hospitals ... has not worked and has been discredited internationally as a failed experiment.

"How many more people must be killed, or hurt ... and how many more reports must the Government receive before action is taken?"

In fact, inquiries generally laid the blame not with the policy but with the performance of mental health services - and not always the result of underfunding. In the case of Mark Burton, who stabbed his mother to death a day after a medical officer released him from Southland Hospital's acute unit, an inquiry exposed a service riddled with organisational failings.

Government financing did improve - and Simpson says the homicide study is one indicator that service levels are approaching what is needed.

"Have we still got gaps out there - yes, we have. Are we closing them at a rate in which the service can actually grow and improve - I think we probably are. Everyone would say there are quality gains to be made, but we have also come a long, long way."

The report offers hope that services are getting it right - most of the time.

"It may be that the improved quality of services that people have striven for in the past decade are preventing more tragedies than we may have expected," he says. "That's the other way of looking at the numbers - we don't know about all the homicides we have saved."

While citing the advent of community support, recovery-based policies, forensic services and better medicines, the study offers pointers for the future. Just over half the mentally ill killers suffered psychotic illnesses such as schizophrenia, conditions associated with delusions and hallucinations that can lead to violence. Half had been admitted to psychiatric hospitals or acute facilities, 10 per cent within the previous month and 20 per cent within the previous year.

But most had been admitted only once or twice in the previous five years and nearly a third had no previous contact with mental health services.

Burdett says although things have improved, some mentally ill people still end up rejecting medication, taking harmful drugs and languishing in boarding houses

The study acknowledges the need to do better in reaching people with psychotic illnesses earlier, including the 30 per cent whose offence is their first episode of illness.

Which sounds fine until you consider that services in Auckland are still stretched to the limit and said to be counting down to the next tragedy. Last year, a crisis team in West Auckland took industrial action against being overloaded with patients, leaving an answerphone message saying "temporarily unable to provide a service due to lack of resources".

In March, crisis workers complained that staff numbers were dangerously low and that bed shortages were forcing them to put patients in police cells. Their workload was forcing them to skimp on patient risk assessments, a practice criticised after the Malcolm Beggs killing.

But Tony O'Brien, crisis intervention worker at Auckland Hospital's emergency department, says the chances of someone not receiving follow-up care are lessening. "It's still difficult to access inpatient beds and acute care. If somebody went into crisis today it would be difficult to get a bed.

"There's always the potential for things not to work for somebody, but I actually think we are doing a reasonable job."

Does this mean society has nothing to fear from mental illness? Not quite, says Simpson. The mentally ill whose illnesses are not well managed do have a greater tendency to violence than the rest of the population.

The second most common diagnosis for mentally ill killers in the New Zealand study was major depression, a common cause of infanticide - which helps to explain the higher proportion of mentally ill women who murder than in "mentally-well" homicides.

"If we can decrease the number of people with active psychotic illnesses or provide better care for postnatal depression - these interventions may lead us to reduce homicides even further."

But Simpson anticipates criticism of the data from both the "reinstitution" lobby and relatives of homicide victims. Given the number of prisoners diagnosed with mental illness, it's obvious that some mentally ill killers have been "missed" by the judicial system.

The report admits the classification of 81 murder suicides as "mentally normal" is arguable, but records indicate that only nine had evidence of previous mental health contact. "Murder suicides were generally domestic tragedies with evidence of immense turmoil and distress, but uncommonly with mental illness."

Simpson says the statistics also highlight the hairline judgments which must be made on psychiatric risk.

It is estimated that 12,000 people have a psychotic illness at any one time. The research found these responsible for half the murders by the mentally ill, maybe two or three each year. "So we are trying to predict a very rare event, a very shocking and horrible event, but nonetheless very rare."

With modern medicine and treatment it's not the diagnosis that defines risk, but the relapse of illness, he says. "The real challenge is to do better at predicting when individual patients are at risk of lapsing into a dangerous state."

He acknowledges that the public, and especially victims, find it difficult to accept that mentally ill killers do get better.

"They have that image of the horror of what occurred frozen in time."

The reoffending rate among people with a serious mental illness is much lower than with people who are not ill because the motivation is different. "People who offend when they are mentally ill generally believe they are doing the right thing; they are driven to act by a moral imperative. Once their moral judgment is no longer affected by illness they are normal and moral people."

From a nursing perspective, Tony O'Brien says the task of predicting risk is "quite fraught".

"If you are talking to someone and they're hearing voices and you know they're abusing alcohol or methamptheatmines or marijuana, you know there's an increased risk.

"But if they are currently well and not abusing, you may know there's a degree of risk but there's no way of telling. Distant risk is a much harder issue that's not well understood."

But while the mental health service can make further progress in reaching potentially dangerous people and diagnosing risk, it's clear from the study that this will have only a tiny impact on the homicide rate.

More must be done outside the mental health sector to prevent homicides, such as targeting substance abuse and alcohol problems, says O'Brien.

Director of mental health David Chaplow says the report should nevertheless help change the perception that we live in a violent country. Our murder rate is on a par with Britain and Australia and a third that of the United States.

"If you have a fairly stable home environment and you stay at arm's length from illicit drugs and alcohol your chances of dying violently by another's hands are almost zero."

Chaplow says more than 30,000 people are in contact with mental health services over the course of a month and perhaps a dozen will come before the courts. "Should we lock up the 30,000? The study also shows that at least 30 per cent [of mentally ill killers] have never come in contact with mental health services.

"It's really a matter of who we target, and we have a better idea now: people with psychotic illnesses and a past history of risk to themselves or others.

"But that 30 per cent figure means we also need to say to the community, 'If you have a loved one who's acting strangely, who has funny ideas and is frightening you, then get help real quick'."