A year of guidelines: What have we learnt?

To function as a mental health clinician, it is necessary to have an awareness of a chain of evidence and information. First, knowledge and evidence are needed regarding the phenomenology and classification of the distressing experiences and behaviours that we come across and are asked to treat. Second, we need some knowledge of risk factors for these phenomena, for example, socioeconomic factors and individual risk factors which might be modified. Next, clinical trial evidence is vital in assessing the effectiveness of treatments – but this is not enough. The clinical trial and phenomenological evidence has to be gathered together in meta-analyses or, ultimately in integrated, systematic treatment pathways such as those featured in guidelines. Finally, there are situations that are either so complex, rare or unusual that the only source of information regarding how to deal with them comes from the collective experience of clinicians. This of course is the area of single case reports or case series which is such an important part of what is reported in some academic journals.

Australian & New Zealand Journal of Psychiatry (ANZJP) this month has data and debate in all five of these areas, beginning with that of phenomenology and classification. Adopting the Christmas spirit in this December issue of ANZJP, we have given readers a break from discussions pertaining to Diagnostic and Statistical Manual of Mental Disorders 5 (DSM-5) taxonomy and the complexities of mixed mood states. However, this issue does contain an important and interesting review of borderline personality disorder in the elderly (Beatson et al. this issue). Beatson et al. argue that this is a particularly important issue given the ageing of the population and the increasing numbers of elderly people in residential care. It is likely to be particularly problematic if the diagnosis is missed and therefore appropriate treatment is not delivered. To improve this situation, Beatson et al. review the phenomenological differences between borderline personality disorder in the younger population and in old age and propose a screening tool to assist in the timely diagnosis of borderline personality disorder. They note that, compared with younger sufferers, the phenomena of impulsivity and identity disturbance appear to be less in the elderly with borderline personality disorder while ‘depressivity’, emptiness and somatic symptoms are increased.

Cognitive impairment is increasingly recognized as an important part of the phenomenology of mood disorders. Daglas et al. (this issue) present data on cognitive function in first episode mania and succeeded in the difficult task of following up a significant number of patients for a year following a first episode of mania. While they report improvement in a small subset of measures, over a year, most of the measures did not appear to change relative to control subjects. The study therefore suggests relatively poor recovery of cognitive function in the year following a first episode of mania and adds to the growing literature that shows significant ongoing cognitive dysfunction in bipolar disorder.

While some risk factors for mental disorders and distress might be modified in individual treatment plans, others are clearly modifiable only by large-scale policy changes. Endicott et al. (this issue) present data on mental disorders and psychological distress, and their association with various demographic and socio-economic factors across Australia. While 1 person in 10 reported recent psychological distress at a high or very high level, this varied dramatically depending on the socio-economic status of the area (from 16.1% to 6.9%), and of course the greatest distress occurred in the most disadvantaged areas. The study confirms a widely held view that socio-economic status has a huge influence on the rates of mental disorder and adds to data suggesting that this should be a major government policy issue.

Clinical trials in patients with mental disorders are, predictably, an issue of great controversy, particularly since the majority of large pharmaceutical trials are conducted by Big Pharma. This issue is once again taken up in the Journal this month by Jureidini (this issue) commenting on a paper of Mulder et al. (2016), which discussed the implications of the reanalyses of study 329. Jureidini argues that transparency cannot be assured while pharmaceutical companies analyse their own data, and suggests that patient level data should be made available for analysis and reporting. He proposes a system in which legitimate researchers can apply for access to this data, and thereby produce their own analyses and then submit these for publication. Clinicians would then be able to assess the different analyses, compare the different outcomes and assess the rigour of the various approaches – taking into account the possible effects of conflicts of interest any researchers may have. Naturally, this would have significant implications for meta-analyses of data, both standard meta-analyses and network meta-analyses, both of which have been criticized on the basis of the possible bias of selective publication or non-publication of clinical trials (Turner et al., 2008). In an analysis of three sets of guidelines on the treatment of mood disorders, McLaren and Cipriani (this issue) argue for a primacy of network meta-analyses in determining guidelines. No doubt, the availability of patient level data to outside researchers, the issue of non-publication of clinical trials and the relative merits of network meta-analyses in determining clinical guidelines will continue to be debated in future issues of ANZJP.

In producing clinical treatment guidelines, the guideline working groups grapple with many of these issues including, in particular, the quality of the evidence and what can be inferred from clinical trials and meta-analyses. A highlight of ANZJP over the past year has been the publication of several sets of guidelines (Carter et al., 2016; Galletly et al., 2016; Malhi et al., 2015) and this month’s issue features reactions to these guidelines. As noted, McLaren and Cipriani (this issue) argue for greater primacy of network meta-analyses. They also argue for a revised system for rating the quality of the evidence for treatments and suggest that guideline groups should go beyond the simple production of guidelines and put in place systems to monitor and assess the effects of the institution of guidelines. Dargèl et al. (this issue) also review the guidelines and make the important point that a key test for guidelines is that they actually change practice, something that may not necessarily happen.

The effects of guidelines can be assessed indirectly perhaps, using large national prescription databases. Kjosavik et al. (this issue) report such an analysis to determine the average duration of treatment with antidepressants in a sample of 10% of patients receiving pharmaceutical benefits scheme prescriptions. The average duration of treatment with antidepressants was four years. Not surprisingly, this average was higher as age increased. The authors note that recommendations have emphasized the importance of prescribing antidepressants for a sufficient period of time. They point out that these data suggest that antidepressants are frequently not ceased and that recommendations should perhaps emphasize the need to review and consider the cessation of antidepressants on a regular basis.

The case reports in ANZJP this month are, as usual, interesting and informative. The issue of patients who have had major improvements on clozapine having to stop this medication because of complications is important, since such patients may then deteriorate significantly. Goh and John (this issue) report on the continuation of clozapine following pulmonary embolus despite other risk factors for thromboembolism, with a good outcome. These sorts of case reports are vital in guiding clinicians in similar situations.

Finally, the findings of two articles this month might have been considered for inclusion of mood disorder guidelines had they been published at the time. Begg and Murphy (this issue) have produced an extremely useful summary of the content of mental health medications which may relate to various allergies and food sensitivities. Table 1 in this paper might usefully be laminated and placed on the walls in the offices of practitioners – or in the modern world made into an ‘app’.

Sadly, laughter therapy is not included as a recommended treatment in the ANZJP mood disorder guidelines. However, Pan and Yeh (this issue) report on laughter therapy in an elderly patient with treatment-resistant depression and comorbid dyskinesia. The patient attended a laughter therapy club where she received a programme of 50 minutes per day, 7 days a week. After 3 months of this, her mood dramatically improved and her dyskinesia became less troublesome. The case report originates in Taiwan but presumably different methods could be used to induce laughter in different populations. Here, in New Zealand, for instance, patients might be able to watch the Wallabies play rugby for 50 minutes a day and achieve the same effect.