Highlights of this issue

First do no harm (with valproate in people capable of becoming pregnant)

The risks associated with using sodium valproate in pregnancy have been well established for over a decade. Indeed, some campaigners have argued that evidence for its teratogenicity was available from animal trials as early as 1974. These risks are now more clearly defined: 40% of babies exposed to valproate in utero are at risk of developmental disorders and 10% are at risk of birth defects. These harms are therefore predictable and to a large extent could be preventable.

In their editorial – kicking off the December issues of BJPsych – Howes et al (pp. 711–713) recognise the incredible importance of sodium valproate in the treatment of epilepsy, mania and bipolar depression, as well as for relapse prevention in bipolar affective disorder. However, they call for greater action in advising women of childbearing ability about the teratogenic harms. In particular, they point to several initiatives – such as the ‘prevent’ programme – that have been implemented over the past 5 years to either reduce prescriptions of valproate in women capable of becoming pregnant or that attempt to educate women about the harms and provide reliable contraception to these women. Yet, in a recent audit of mental health trusts, only a quarter of women aged under 55 years who were prescribed valproate and for whom pregnancy was biologically possible were offered the full benefits of the ‘prevent’ programme.

With the extensive guidance and information now available, there seems to be little excuse not to adequately counsel and support women considering a valproate prescription. This editorial presents a serious call to arms for clinicians to consider their own prescribing practices, ensure they are aware of regulatory requirements around valproate prescription and always address the risk of harm in anyone for whom pregnancy is biologically possible.

Expanding scope for Clozapine prescribing?

Another psychotropic medication around which there have been longstanding safety concerns is clozapine. Yet, conversely to the above, evidence suggests that clozapine is underused in many settings and is often associated with a long delay before treatment initiation in treatment-resistant schizophrenia. Two papers this month give food for thought about clozapine-prescribing practices going forwards. Butler et al (pp. 740–747) present data from South London showing that community initiation of clozapine was at least as successful as in-patient titration and associated with reduced out-patient visits, reduced psychiatric hospital bed days, reduced service costs and improved symptoms.

In their study, Lähteenvuo and Luykx et al (pp. 758–765) used data from two national cohorts in Finland and Sweden to explore the role of clozapine when used in people with schizophrenia who may also have a substance misuse diagnosis. Clozapine was associated with a reduced risk of developing substance use disorder compared with antipsychotic polytherapy, and clozapine was associated with lower relapse rates in patients with both diagnoses. An important finding given the scale of substance misuse comorbidity in this patient group and the lack of treatment support often practically available.

In a further paper exploring prescribing data in Northern Ireland but specifically concerned with the COVID-19 pandemic, Maguire et al (pp. 748–757) found that the predicted tsunami of antipsychotic prescribing and monitoring during the pandemic did not come to fruition. Interestingly, there was a slight uptick early in the pandemic, when people perhaps decided to stockpile their medications alongside toilet roll and pasta.

In other studies this month

Cui et al (pp. 732–739) describe how deep learning methods are employed to identify differences in grey and white matter and demonstrate multidimensional neuroanatomical changes in schizophrenia that can robustly discriminate between people with schizophrenia and healthy controls. Importantly the authors recruited from multiple sites across China to try to capture the heterogenity in symptoms and scanners that occurs in ‘real world’ clinical environments.

Despite concerns about the cognitive impact of psychosis in the early years following diagnosis, a meta-analysis by Watson et al (pp. 714–721) of 25 longitudinal studies finds no evidence of cognitive decline or improvement during this period. However, the patient groups were impaired on all cognitive domains at baseline, supporting previous findings that much of the cognitive impairment seen in psychosis seems to occur prior to onset of illness and remains relatively stable after symptom onset.

Finally this month is a paper exploring transdiagnostic categorisation of the symptoms of mania and/or irritable episode symptoms. Arathimos et al (pp. 722–731) demonstrate diagnostic overlap, genetic risk overlap and overlaps in functional outcomes with these symptoms. They believe that an improved classification system for psychopathology could consider a weighted approach to symptoms and thus facilitate a move ‘toward a more dimensional classification of mood disorders’.