Categories
l

NHS Talking Therapies Proceeds by Copying

but as anyone knows who has copied a fellow pupil’s homework, this can end in tears! The Service believes, it disseminates ‘best practice’ by networking clinicians together. But there is no evidence that this is the case.Those most vocal or most eminent are likely to hold sway. Ensuring the dominance of eminence not evidence-based treatment. This runs the risk of creating a totalitarian culture in which it is not possible to express a dissenting opinion. The power-holders then control what is published in journals/publications. But like the Russian regime of the 1980’s they would vigorously protest that they are engaged in Glasnost (0penness).

The use of artificial intelligence in mental health is an exemplar, ‘par  excellence’ of copying. The authors of IESO claim that its interactive text messaging service for clients, is based on the careful selection and analysis of client-therapist exchanges and that they have determined what works. If indeed this were the case it would make the provision of mental heath services much cheaper for the NHS (as well as making a nice profit for IESO). But no such reliable database with real-world outcomes e.g being free of a disorder as assessed by an independent observer for at least 8 weeks, has been determined.There are no published papers to substantiate a real-world effect. Without fidelity checks there can be no assurance of the translation of an evidence-based treatment to routine practice.

I recently read James Davies’s book Sedated: How Capitalism Has Created the Mental Health Crisis, (2021) Atlantic Books, without necessarily buying into his thesis that Capitalism is the arch-enemy, it is certainly the case that financial factors ( NHS Talking Therapies cost the taxpayer £ 2 billion a year), coupled with the opportunity to exercise unbridled power has produced a dysfunctional mental health system. But the problems are not confined to primary care, I recently saw a client who has had 4 years of secondary care treatment in relation to child abuse, the records revealed no objective audit. This echoes the National Audit Offices failure to provide any credible independent assessment of mental health services. There is an inherent believe that the mental health services must be all too the good because the clinicians are well-intentioned. But if they had stopped long enough to listen to my client he would have told them ‘I need a new approach, I have plateaued’. In reality there is no Glasnost. Let’s no be copy-cats.

 

Dr Mike Scott

Categories
l

NHS Talking Therapies Digital Courtship

 

 

Tomorrow the Service is hosting a 3-hour webinar funded by two artificial intelligence agencies IESO and Limbic. This comes hot on the heels of the Irish Advertising Standards Agency finding SilverCloud guilty of breaching advertising standards in citing its’ recovery rate. In fairness I could see no such extravagant claims from IESO.  Nevertheless, IESO claims its:

 

 

‘app-based programme is comprised of a unique blend of evidence-based CBT approaches. They have been carefully selected having been shown, through analysis of ieso’s dataset of 750,000 hours of delivered therapy, to be the most effective with patients’.

 

But there is no published evidence that this ‘unique blend is effective’. Nor is there any evidence that the 750,000 hours of delivered therapy were assessed independently with fidelity checks to ensure that the alleged treatments actually took place.   

 

There is a computer adage GIGO – ‘garbage in, garbage out’ . There is no assurance  that the programme does not constitute ‘garbage in’.  But the programme is it seems user-friendly with a conversational style aping programmes for building your own website.  Unsurprisingly it has therefore gained traction:

Through our award-winning digital platform, ieso has delivered therapeutic interventions to 135,000+ people with over 750,000 hours of treatment provided. We use a typed modality which captures, with permission, the exchange between a therapist and patient all in an ISO27001-compliant manner’. 

There is an aura of misleading scientific respectability, is this yet a further breach of advertising standards?  AI may have a potential impact, but this offering is for the forseeable future likely to frustrate the consumer, who at a minimum will find it too            time-consuming.

 

Dr Mike Scott

 

Categories
l

NHS Talking Therapies’ Partner SilverCloud Found Guilty of Breaching Advertising Standards

 

The Irish Advertising Standard Authority, concluded  on  October 3rd 2024, that SilverClouds advertisement of a recovery rate “up to 65%” should not be published again.  In their judgement the IASA said  that no evidence had been provided to substantiate the recovery rate and expressed concern at the advertiser’s failure to respond.  Yet SilverCloud claims to support “80% of NHS Services”. Both NHS Talking Therapies and SilverCloud have refused to engage in any public debate. Despite attracting Government funding of £2billion a year, there has been no publicly funded independent evaluation of effectiveness. 

 The best independent evidence Scott (2018), suggests that only the tip of the iceberg recover.  Unfortunately NHS Talking Therapies, are it seems answerable to no one and there is no indication that they are abandoning their absurd claim of a 50% recovery rate.

The claim that 65% achieve clinically significant improvement, has a scientific aura about it for a member of the public, but with no idea as to what it means in real world terms. They might consider that as ‘77% of volunteers agreed that volunteering improved their mental health and wellbeing’, National Council for Voluntary Organisations survey 2019, this was actually a better investment.  The implication of SilverCloud’s claim is that it is the use of their product that has brought about the ‘significant improvement’ but this has not been demonstrated. The said improvement could represent regression to the mean, patients presenting at their worst and becoming a bit better with time. Equally it could represent the client’s wish to please the therapist and/or to feel that they have not wasted their time. The claim of ‘significant improvement’ because of usage of the product is not evidence-based and is simply a marketing ploy.

NHS Talking Therapies have produced no evidence that its’ therapists using SilverCloud make any added difference to their clients over and above that of those who didn’t use it. But the juggernaut roles on, with SilverCloud advocated by NHS Scotland for those with long-term conditions!

Dr Mike Scott

Categories
l

NHS Talking Therapies – No Better Than Placebo

NHS Talking Therapies boasts a 50% recovery rate. This doesn’t square, easily with the Prime Minister’s view that the NHS is ‘broken’.

A letter from Lord Darzi to the Secretary of State for Health and Social Care, dated 25 September 2024 states 

 ‘The first step to rebuilding public trust and confidence in the NHS is to be completely honest about where it stands’. But nobody is telling the truth, that NHS Talking Therapies is performing no better than placebo. Lord Darzi in paragraph 18 of his report      re-iterates NHS Talking Therapies claim of 50% recovery, with approval and without any quibble.

Marketing its’ mantra of a 50% recovery, NHS Talking Therapy has secured Government funding of £2billion a year. [Effect Size (ES) is a measure of change with an intervention]. But the published NHS Talking Therapies results show that the Effect Sizes ( i.e the difference between the post and pre-treatment scores divided by the pooled standard deviation) are no different to a placebo response. The placebo response takes account that disorders take an episodic course (i.e that symptoms naturally vary over time), and that people respond to attention, compassionate care and regression to the mean (improvement over time, first attending at their worst). Further it makes no difference whether the Service delivers alleged CBT or anything else, casting doubt on the meaningfulness of its training regime.

In the table below the Effect Size in NHS Talking Therapies for CBT (Table 4n, Annual Report 2022=2023) is contrasted with the Effect Size for placebos from the Bschor et al (2024) study.

[Extract from  Bschor et al (2024) Figure 1. Random-Effects Meta-Analysis Estimates of Pooled Pre-Post Placebo Effect Sizes ]

Diagnosis

Effect Size (Placebo)

Bschor et al (2024

CBT Effect Sizes NHS Talking Therapies PHQ9 (GAD7)

MDD

1.40

1.0 (0.8)

GAD

1.23

0.8 (1.2)

Panic disorder

0.92

0.7 (1.0)

PTSD

0.84

0.8 (0.9)

Social phobia

0.72

0.8 (1.0)

 

NHS Talking Therapies uses a category of ‘Anxiety’ for the biggest grouping of those receiving CBT with an ES of 0.7 on the PHQ-9, and 1.0 on the GAD-7. They thus do no better than placebo. The Service also uses a category ‘Mixed Anxiety and Depressive Disorder’ , with respective Effect Sizes of 1 (PHQ-9) and 1.1 (GAD-7), again no better than placebo.  But ‘Anxiety’ and Mixed Anxiety and Depressive Disorder are not categories used in the Bschor et al (2024)study or in DSM-5-TR. Not only does NHS Talking Therapies CBT appear ineffective but the Service uses its’ own nomenclature, making comparison’s difficult.

NHS Talking Therapies, for anxiety and depression, Annual reports, 2022-23

However, NHS Talking Therapies Categorises Its’ Treatments, There Is No Discernible Effect On Outcome

The results are no different if just those at caseness (PHQ-9 score  10 or more) are considered, see table below:

   

No 0f cases

633236

Start        

sd

End

sd

ES

Dep

 

268743

17

4.7

10.2

6.5

1.5

GAD

 

174049

13.4

5.5

8.1

5.9

1

PTSD

 

35234

17

5.5

11.3

7.2

1

Mixed

 

16677

16.1

5.1

10.1

6.5

1.2

Socl

 

18320

13.9

5.7

8.7

6.2

0.9

Panic

 

14759

13.8

6

8.5

6.5

0.9

OCD

 

16105

13.4

6

8.5

6.3

0.8

Hyp

 

9316

11.4

6.1

6.2

5.7

0.9

Agor

 

4487

15

6.1

10.4

7.1

0.8

Specific

 

4717

11.5

5.8

6.7

5.7

0.8

 

Thus, it appears to make no difference to whether NHS Talking Therapies delivers alleged CBT, or whatever it delivers. Casting doubt on the meaningfulness of its’ training regime.

 
   

Table 4n Cognitive behaviour therapy (CBT) outcome scores means, standard deviations, effect sizes and outcome measures by diagnosis and organisation, 2022-23

NHS Talking Therapies, for anxiety and depression, Annual reports, 2022-23

 

Dr Mike Scott