More than 10 million people are currently in prison worldwide, and substantially larger numbers of ex-prisoners are living in society. Despite reported decreases in violence in many countries, repeat offending remains high across many high-income and middle-income countries. In the USA and UK, more than a third of released prisoners are reconvicted for a new crime within 2 years, and more than half within 5 years. Furthermore, about 70% of those convicted in the USA are repeat offenders. In England and Wales, this figure is estimated at 90%, and the proportion of individuals convicted who have had 15 or more previous offences has been increasing since 2008.

So begins a new paper that considers Swedish prisoners and the likelihood of reoffending with a particular focus on those with mental disorders. This opening paragraph well describes our incredible success at crime reduction – historical lows across the west – and our incredible failure – those who do end up in jail often are convicted again (and again and again).

This week’s Reading: “Psychiatric disorders and violent reoffending: a national cohort study of convicted prisoners in Sweden” by Zheng Chang et al., which was just published in The Lancet Psychiatry.

Zheng Chang

This paper may not seem immediately relevant – it draws on European data and was published in a British journal. In fact, it is. First, Canada’s prison population has a high rate of mental illness (an International Journal of Law and Psychiatry study found that a third of people entering Canadian penitentiaries met criteria for mental illness.). Second, there is growing interest in prisoners and mental health, both in Ottawa and beyond. Chang et al. helps shed light on this important issue – and perhaps even points the way to more sensible public policy.

You can find the paper here:

Here’s what they did:

· In a longitudinal cohort study, the authors looked at Swedish prisoners who were imprisoned and released within a ten-year period (January 1, 2000 to December 31, 2009). They investigate the sub-population with psychiatric disorders.

· They drew on various Swedish databases including: the National Crime Register (information on all criminal convictions), the National Patient Register (diagnoses for all inpatient and outpatient care), and the Multi-Generation Register (information on biological relationships for all individuals living in Sweden).

· Psychiatric diagnoses were based on the National Patient Register before release and the authors considered co-morbidities (including substance).

· Violent crime was used as a measure, and included homicide, assault, robbery, arson, any sexual offence (rape, sexual coercion, child molestation, indecent exposure, or sexual harassment), illegal threats, and intimidation.

· Using different statistical tools, the authors attempted to take into account demographics and age. They also considered genetic factors (a first-of-its-kind analysis looking at twins). They used the Population Attributable Fraction to quantify the proportion of violent reoffending related to mental illness.

Here’s what they found:

· 47,326 prisoners during the study period; they were overwhelmingly men (43,840 male and 3,486 female prisoners).

· For male prisoners: 18,563 of 43,840 (42%) had been diagnosed with at least one psychiatric disorder before release, and 10,884 (25%) reoffended for violent crimes during follow-up. For female prisoners: 2,233 of 3,486, or 64% had been diagnosed with psychiatric disorder, higher than had male prisoners, but fewer (379 or 11%) reoffended.

· The most common category of crime: assault (or 64% of individuals reoffending), followed by threats and intimidation, robbery, sexual offences, and homicide.

· “Prisoners with any psychiatric disorder had a higher rate of violent reoffending than did those without a disorder.”

· Median time to reoffend was 2-4 months shorter in men with psychiatric disorder than those without, and 4-8 months shorter for female prisoners than those without.

· Prisoners with psychiatric disorder had a high probability of violent reoffending: over 5 years, the probability was 0.41 for male prisoners with psychiatric disorder and 0.25 for those without. In female prisoners: 0.2 and 0.08. (See the graph below.)

· 10,884 incidents of violent reoffending occurred in male prisoners after release; 2,187 were potentially attributable to psychiatric disorder – or a PAF of 20%. This corresponds to a PAF of 20%. The female prisoner group had lower incidents of violent reoffending (379), but a PAF of 40.

The authors write:

In this longitudinal study, we have shown that psychiatric disorders were associated with a substantially increased hazard of violent reoffending. The association was independent of a number of measured sociodemographic, criminological, and familial factors, except that the finding in female prisoners was non-significant when taking familial factors into account. To our knowledge, we are the first to use a sibling design to study reoffending in an unselected prison population.

The authors conclude:

Many individuals with psychiatric disorders revolve between admission to hospital, homelessness, and the criminal justice system. Our findings underscore the need for improved detection, treatment, and management of prisoners with mental health disorders, and linkage of these prisoners to community-based mental care services on release. They also emphasise the need for further research into the role of psychiatric diagnoses in risk assessment for future offences and the effectiveness of diversion from criminal justice.

A few thoughts:

1. This is a carefully and thoughtfully constructed study.

2. That said, there are clear limitations here. Start with the fact that psychiatric disorders were drawn from patient registers; as the authors point out, this database probably underestimates the prevalence of mental illness. Another limitation: the dataset is Swedish. Now, for the record, I wish to state that some of my best friends are Swedish and that the Swedish chef was the stand-out talent on the original Muppets series (a comic genius, and clearly more talented than the much over-rated Kermit The Frog). That said, the Swedish prison population tends to be somewhat different than that of other countries and the Swedish prison system on the whole is different.

3. The conclusion has broad public-policy implications. Historically, prison psychiatry has been a low priority. Should this be reexamined for public safety?

4. I also note that across western countries, political discussions of crime tend to be superficial. Right-leaning politicians tout tough-on-crime measures that ignore falling rates of crime; left-leaning politicians tend to highlight societal woes like poverty as a source of crime that ignores the fact that re-offending is the big problem. This study suggests that a more focused approach on prisoners with mental health may reduce crime further.

Reading of the Week. Every week I pick a reading — often an article or a paper — from the world of Psychiatry.