Thinking around suicide

Is sensible thinking finally coming to address the question of suicide prevention? Jorm (this issue; Figure 1) nicely summarises the impact of various interventions and plans introduced to help prevent suicide in Australia since 1991; there is no discernable impact. A recent letter to the Lancet notes that prevalence of antidepressant use in Australia rose from 1.4% in 1993 to 10.7% in 2016/2017, while mental health treatment through Better Access rose from 0.8% in 2006/2007 to 4.7% in 2016/2017 (Bastiampillai et al., 2019) again with no change in suicide rates over this period. Suicide is a tragedy on many levels; so, this lack of success in reducing rates should, as Jorm notes, lead us to ask what we could do differently. Responses will follow and we welcome readers’ views.

Sher (this issue) helps begin the debate. As he points out, there are two main preventive strategies in public health: the high-risk approach and the population approach. In the field of suicide prevention, we have focused on the former; all psychiatric disorders are associated with an elevated suicide risk and some patients have a higher risk than others. There are at least two problems with this strategy: we are poor at accurately determining which individuals are at a high risk of suicide, and so, this strategy is not working. Sher suggests that a population-orientated approach might be beneficial. The effect of reducing a relatively small suicide risk for a large population may be more effective than reducing the risk among a small group of high-risk individuals. Clinically, this suggests that treating all patients well rather than focusing resources on so-called high-risk patients who we cannot identify accurately in the first place is a better model for clinical practice.

Reviewing clinical practice

Two systematic reviews have possible clinical implications. Fairbrother et al. (this issue) point out that while we are all aware that lithium treatment can cause hypothyroidism, there is also evidence that it may cause hyperthyroidism. Although an uncommon side effect of lithium, it is still higher than that found in the general population. As the authors note, the quality of the studies is not high, but there are possible clinical implications. Hyperthyroidism may mimic symptoms of an affective disorder and may alter the renal clearance of lithium leading to toxicity. These findings reinforce the need for regular thyroid monitoring in patients taking lithium, particularly if there are changes in their mental state.

The meta-analysis by Myles et al. (this issue) reports the surprising finding that data from controlled trials do not support the clinical impression that clozapine has a stronger association with neutropenia than other antipsychotic medicines. Moreover, the strength of the pooled association is weak, casting doubt on whether clozapine is causally associated with neutropenia. As the authors note, this challenges the assumption that clozapine has specific casual and clinically important haematological risks that do not apply to other antipsychotic medications. The finding also implies that either all psychotropic drugs should be subjected to haematological monitoring or, alternatively, none of them should be. Will this change practice? It might in the future, but probably not yet. Again, as the authors point out, the rarity of events such as neutropenia limits the power to detect differences in rates between drugs, and even a meta-analysis does not conclusively prove the absence of a small, but specific, association between clozapine and neutropenia. However, the findings suggest that future research using large databases is warranted and might eventually lead to increased ease of prescribing for the most effective treatment for schizophrenia to date.

Imaging

Two imaging studies are on very different populations. Both are quite large for these types of studies, giving some confidence in their findings. Delvecchio et al. (this issue) report different localization of grey matter reductions between elderly bipolar disorder and frontotemporal dementia, which they suggest supports their discrimination as distinct clinical entities. Malhi et al. (this issue) studied the effects of childhood trauma on the hippocampus in adolescent girls. The findings are complicated. These were higher emotional trauma–induced volumetric changes in the left CA3 hippocampal subfield volume, but varied depending on age. They proposed that changes to the integrity of hippocampal subfields in response to increasing emotional trauma exposure may provide a mechanism through which vulnerability to mood disorders may be increased in adolescent girls.

Other things

Prescribing practices of Indian psychiatrists in bipolar disorders (Reddy et al., this issue) suggest broad adherence to the recommendations of clinical practice guidelines, although antidepressants may be overprescribed and maintenance treatment only prescribed for a limited period. This should be seen in the context of 5000 psychiatrists treating over 1.3 billion people, that is, one psychiatrist per 260,000 population. That is the equivalent of 20 psychiatrists for the whole of Sydney – or just under two psychiatrists for the whole of Christchurch. Perhaps, some allowances should be made.

Judd et al. (this issue) report on the non-specificity of the Edinburgh Post Natal Depression Scale (EPDS). The 43% who scored 13 or more on the EPDS were more likely to have borderline personality disorders (BPDs) or borderline personality traits than depressive or anxiety disorders. As the authors note, BPD is associated with being a ‘high risk’ caregiver and interventions are indicated. They suggest patients with high EPDI scores should be seen by a clinician who looks beyond depression and anxiety.

Vine et al. (this issue) report that reformed mental health legislation introduced in 2014 has been associated with shorter community treatment orders (CTOs) and a reduction in CTOs in the 2 years following the index discharge. However, there was an increase in the return to inpatient orders. As Ryan (this issue) and Vine et al. note, these changes may be due to legislative reforms or numerous other factors, including a depletion of funding in the Victorian mental health sector. Ryan remains concerned and presents some evidence that the reforms are not yet known or understood by many psychiatrists, so the impact on patients may be minimal. In particular, the need to accurately assess the decision-making capacity of people with mental illness may be less than ideal. So, a varied diet in this month’s issue, but considerable food for thought.