Niki Ellis | Saddling unicorns


Saddling unicorns

05 Sep 2017, Posted by Professor Niki Ellis in Inside OHS articles

“When mental [illness] increases until it reaches the danger point, do not exhaust yourself by efforts to trace back to original causes. Better accept them as inevitable and save your strength to fight against the effects.”


I was amazed to find this quote on the ‘interweb’. It was attributed to George Sand, the famous French novelist, who was a woman, born Aurore Dupin, in 1804 and who died in 1876.

I found it shortly after reading an article by Nick Haslam. No, not the ageing British interior designer you read about in Tatler, that’s Nicky Haslam; this Nick Haslam is a psychology professor writing for The Conversation.

Haslam said that historically whilst psychiatrists and psychologists tend to favour different root causes, the former neural, the latter cognitive malfunctions, they share ‘the idea that a cluster of symptoms can be traced back to an underlying pathology’.

Haslam argues that this is a fruitless exercise as the relationship between symptoms and causes are not unique.


Quixotic searching for an underlying cause for mental illness


He uses a botanical metaphor, bamboo, to illustrate this.

In this plant ‘an interconnected network of underground roots (hidden causes) generates many visible stems (symptoms). No stem can be traced back to a single root, and no root feeds a single stem’.

Haslam describes a new conceptualisation of mental illness in which networks of symptoms are mapped and central symptoms “related to many others” are identified and given priority in treatment.

The article is worded quite strongly, the search for an underlying cause is “quixotic” or “like trying to saddle a unicorn”, and “treatments should directly target particular symptoms, not a fictitious hidden cause”.

Three months earlier another article by Ashley McAllister, Maree Hackett and Stephen Leeder flew in on Twitter.

This reported on a literature review which aimed to analyse how disability income support schemes in Australia and Ontario determined eligibility for mental illnesses.

The paper noted the challenges of mental illnesses for such schemes, compared to physical illness, including fluctuations, lack of objective diagnostic criteria and lack of objective symptoms.

The study concluded “that disability income support, especially the assessment process, is not adequately designed for mental illnesses”.

The authors went further saying the “policy tools, ie the Impairment Tables … are inadequate to assist in interpreting the definition of disability”.

The consequence of this was that decision-making in the schemes relied largely on the judgement of the individual assessors.

So if Nick Haslam is right about mental illness (Nicky Haslam is always right about the colour and proportions of the decadent but liveable spaces he designs for his celebrity and aristocratic clients), the solution to the problem described by McAllister, Hackett and Leeder, may be for income support schemes to focus more on symptoms and less on diagnoses.

And to some extent this is starting to happen.

Sort of. Big data is allowing improvements in the use of predictive analytics in the personal injury insurance sector.


Triaging claims using algorithms


It is best practice now to triage claims, using algorithms to assess risk of delayed recovery, and to stream claims into services appropriate for their level of risk.

When this practice began, its success was limited by the data used in the models developed to predict risk of a delayed recovery. They were based on data readily available to the schemes such as claims data and diagnosis.

Yet we now know (and in fact have known for quite a long time) the risk factors important for delayed recovery, and therefore time on income support, include psychosocial factors, ie as much related to the characteristics of the person and their circumstances, as their diagnosis, especially for common disorders such as musculoskeletal and mental illness.

At the moment we are seeing schemes adding psychosocial data into their triaging models. We can expect that in the future predictive analytics will be extended to establish expected milestones in recovery pathways.


Provide evidence based tools to case managers


Once these are in place, the system will alert case managers to those people who are not progressing as would be expected and may need additional help.

Ideally at this point (sometimes referred to as re-triaging), the system would provide evidence-based decision support tools to the case manager, or perhaps the system itself would automatically ‘decide’ and refer to appropriate services.

Previously, schemes were beholden to the bio-medical model and hostage to its limitations, oh so apparent in primary and secondary psychological ‘injury’.

Now we are seeing a broader, bio-psycho-social view. Treatments and services will change too to reflect this.

After all, if we identify psychosocial factors as reasons for a slower than expected recovery, we will need services to address them.

I can see similarities to Nick Haslam’s concept of focussing on symptoms and outcomes rather than trying to saddle the unicorn of diagnosis.

So if I could invite anyone, alive or dead, to dinner to talk about the treatment of, and income and other support for mental illness, I would include George Sand, Nick Haslam, Ashley McAllister, Maree Hackett and Stephen Leeder.


The dinner would be held in a Nicky Haslam designed dining room, and the guests would arrive on unicorns.




First published in Thomson Reuters Inside OHS, 01/09/2017

Inside OHS Editor: Stephanie D’Souza; (02) 8587 7684; Stephanie.D’