Pharmacotherapy for incarcerated people with a history of violence: Response to commentary by Schofield et al.

In their recent article, Schofield et al. (2019) discussed the need for placebo-controlled trials of medication treatments to prevent further violence among people in prison. They identified men convicted of a violent offence who scored high on an impulsiveness scale and found that many of them would be willing to participate in this research.

Effective, uninterrupted pharmacotherapy is likely a key component of improving health and reducing recidivism among people with mental disorders released from prison. Schofield et al. suggest the selective serotonin reuptake inhibitor (SSRI) sertraline as a suitable medication to prevent violence. While the neurobiology of aggression/violent behaviour involves many different neurotransmitters, low levels of the serotonin metabolite 5-hydroxyindoleacetic acid (5-HIAA) have been positively associated with violent behaviour among rodents and humans, providing a rationale for using SSRIs in this context. However, evidence for treating violent behaviour with SSRIs is far from clear. Whole-population cohort (Molero et al., 2015) and clinical trial data in healthy volunteers (Bielefeldt et al., 2016) suggest SSRIs are associated with increased risk of violence. This risk appears particularly heightened in young men (Molero et al., 2015) – a group overrepresented among people with a history of violent offences.

As Schofield et al. point out, people with a history of violence, including those in prison, have a high prevalence of multiple mental disorders. Many of these disorders have established, evidence-based pharmacological treatments: notably, opioid substitution therapy for opioid use disorder; psychostimulant therapies for attention-deficit hyperactivity disorder; and antipsychotic medication for psychotic disorders. Effectively treating mental disorder among people in, or released from, prison does not guarantee improved health or a reduction in violent outcomes. Nonetheless, a large cohort study of people released from prison in Sweden showed addiction treatment drugs, psychostimulants and antipsychotics were associated with lower rates of violent re-offending, while antidepressants were not (Chang et al., 2016).

Custody provides an ideal setting to conduct a mental health assessment, identify coexisting physical and social needs and start pharmacotherapy where indicated. Most episodes of incarceration are short; therefore, integration with community health providers to ensure continuity of care after release from prison is critical. This period – particularly the first 30 days – carries a high risk of preventable morbidity and mortality (Spittal et al., 2019). However, continuation of prison-initiated pharmacotherapy is typically low. This may create problems for people prescribed an SSRI, because withdrawal syndromes are common and can occasionally be severe. Treatment interruption or short durations of SSRIs have been associated with serious adverse events including aggressive/violent behaviour and increased odds of violent suicide.

Social determinants of health also affect medication continuity for people released from prison. For example, stable housing has been associated with improved retention on pharmacotherapy among previously homeless individuals. Among people released from prison, pharmacotherapy is unlikely to be effective without support to obtain housing and employment and to engage with community mental healthcare and alcohol and other drug treatment. Unfortunately, there are many structural barriers to care within prison and after release. One avoidable, yet prominent, structural barrier to continuity of pharmacotherapy in Australia is the exclusion from Medicare, and thus the Pharmaceutical Benefit Scheme (PBS), for people in custody. Furthermore, health services in prison, including forensic psychiatry services, face increasing challenges in meeting the needs of a rapidly expanding prison population. Equitable access to treatment across populations, in particular to ensure that Indigenous people are not further disadvantaged by not being offered culturally appropriate treatments, is vital.

As Schofield et al. note, well-designed violence prevention studies targeting people released from prison are needed. However, trials examining the effectiveness of SSRI treatment have often had substantial bias and attrition and have not examined long-term, ‘real’-world outcomes. The shared neurobiology of violent behaviour and SSRIs and a lack of knowledge of the long-term outcomes of SSRI pharmacotherapy among people released from prison provide a preliminary argument for a trial. However, given evidence for increased serious adverse events from SSRI treatment, the challenge would be to carefully design a trial which is transparent, independent and adhere to the principles of ethical conduct, including clinical equipoise. Given the capacity to link administrative health (including medication dispensing) and justice data in both Australia and New Zealand, such a trial should complement information gathered through direct contact with participants, with linked administrative records to evaluate the effect of SSRI treatment and pharmacotherapy adherence on both symptom profile and long-term ‘real’-world health, violent and justice outcomes among people released from prison.

As Schofield et al. have argued, more resources directed at preventing violence among people released from prison are urgently needed. However, before turning to treatments without an evidence base such as SSRIs, the priority should be to identify people in prison with mental illness and/or substance use disorder, initiate evidence-based treatment and ensure continuity of care after release from prison.