Michael J Scott

  • It is, as if NHS Talking Therapies is housed in a Georgian townhouse. It is not fit for those with a ‘mental disability’. Following a telephone conversation, with a receptionist, you may arrive at the address at the appointed time, only to find your path is barred by a series of steps. Given your lack of ‘mobility’ you may elect to go home. A third of people (33.1%) did that in the year ending March 2025. NHS Talking Therapies Annual Report.

    Once inside, you are greeted by a receptionist who clerks you in, takes your biographic details and gives you two questionnaires to complete. The PHQ-9, a measure of the severity of depression and the GAD-7, a measure of the severity of anxiety. The receptionist might decide that you are not suitable for the service because of your drinking or alternatively that because you are suffering from PTSD/social anxiety disorder/body dysmorphic disorder, you should go straight to the 1st floor where the High Intensity Therapists are housed. Low intensity CBT therapists occupy the ground floor, offering usually up to 6 sessions.

    Almost half of those who enter the building (43.8%) do not complete treatment (defined by the Service as attending 2 or more sessions) according to the latest NHS Talking Therapies for anxiety and depression Annual Report (September 2025).

    Arbitrary Signposting and Treatment

    But there is no transparency about the receptionist’s directions, no specification as to how much drinking is acceptable nor the criteria used to decide on PTSD, body dysmorphic disorder or social anxiety disorder. The receptionist’s lanyard may tell you that he/she is a Psychological Wellbeing Practitioner (PWP) or en route to be a PWP. A brief search on Google reveals that PWP’s are not trained therapists. The PWP has power but lacks credentials.

    The NHS Talking Therapies clinicians are not trained in diagnosis, but the doors on both floors each bear the label of a disorder.  The myth is that behind each door a therapist is compliant with delivering the designated disorder. But no fidelity checks have ever been conducted.

    If initial ministrations are deemed unsuccessful (PHQ-9 or GAD-7 scores above the thresholds of 10 and 8 at the end of treatment) they are sent upstairs to the Hi-intensity therapists. About 10% of clients climb the stairs to the 1st floor but this climb typically takes weeks. 

    Thanks, But No Thanks

    In 2022-2023, for the most common disorders depression and GAD almost as many people attended just one session 235,701 as attended two or more sessions 442,792. Scott (2024).

    The one session attenders appear no different to those who attend two or more sessions, based on psychometric test scores. Yet the former are half as numerous as the latter, but have been ignored by the Services researchers. I had to get details on them from a Freedom of Information request Scott (2024).

    The building is not fit for purpose. But the rent costs about £1 billion a year, with similar expenditure for the nearby Child and Adolescent Mental Health building. A radical re-think is necessary, but the NHS Talking Therapies juggernaut shows no signs of pausing, might this change when it becomes part of the Department of Health?

    Dr Mike Scott

  • The Service was set up to address cases of depression and anxiety,  with a mirroring of treatment for these disorders in the CBT randomised controlled trials (RCTs). In the RCT’s there was an average recovery rate of 50%. But these trials excluded patients with a personality disorder. Given that personality disordered clients are commonplace in NHS Talking Therapies, it is scarcely credible that the Service should achieve a claimed comparable recovery rate.

    In a representative sample of IAPT clients, Hepgul et al (2016) found that 69% had a high risk for a personality disorder with 16% meeting criteria for borderline personality disorder. However, NHS Talking Therapy clinicians do not have the skills to either identify or treat personality disordered clients. Their clinicians literally do not know what they are dealing with.

    Stepped-Care More Apparent Than Real

    The intent was that NHS Talking Therapies would focus on straightforward cases of anxiety and depression, with severe pathology: personality disorders, bipolar and psychosis, managed by secondary care. But it hasn’t quite worked out that way. As many as 35% of IAPT clients (the predecessor of NHS Talking Therapies) exhibit clinically significant psychotic experiences [Perez et al (2017) and 61% scoring above the screening threshold for bipolar disorder Knight et al (2020). Those scoring highly on a Psychotic Experience scale had a lower recovery rate. But NHS Talking Therapie’s gatekeepers, Psychological Wellbeing Practitioners do not have the skills to identify, much less treat severe mental illness.

    Within NHS Talking Therapies cases of PTSD, social anxiety disorder and body dysmorphic disorder are supposed to bypass low intensity and go straight to high intensity. But this presupposes that the PWPs can reliably identify them. However there is nothing in their typically 45 minutes telephone assessment that is a guarantor of reliable assessment of these, or indeed any disorder. The stepping up is again more apparent than real.

    In mental health, stepped care is an attempted replication of what occurs in physical health, where therapeutic resources are maximised by using the least costly intervention first, proceeding to more invasive/costly intervention if the cheaper option does not work. But in physical health there is more reliable diagnoses.Whereas in routine mental health scarce attention is given to diagnosis and for many practitioners it is anathema. With no agreement on what ‘works’, it is welcome to ‘Bedlam’.

    Zavlis (2023) has observed:

    ‘UK national data suggest that: (1) of the 1,647,716 IAPT referrals in 2019/20, 63.21% did not complete treatment and (2) of those who completed treatment, around 60% did not achieve clinical recovery. [ Using the Service’s own metric – my Comment] Although speculative, it may not be farfetched to assume that many such cases of treatment resistance are due to comorbid personality difficulties (among other comorbidities, of course)’.

    Complexity Bias

    The ‘get out of jail card’ for psychological therapists is that their client is ‘complex’. Service providers too, can easily appeal to ‘complexity’ to cover poor performance. ‘Complexity’ has become a selling point for workshops and books. However it is chameleon-like, with no agreed and consistently applied definition. The American Psychiatric Association has, in my view, rightly refused to recognise ‘complex PTSD’. A complexity bias operates in UK mental health:

    Starting in 2009 I wrote a trilogy of books under the ‘Simply Effective’ title published by Routledge. The starting point was simple, an open-ended interview in which the client has the space to tell their story. Followed by questions about each of the symptoms that comprised, possible diagnoses. With treatment targets pertinent to the diagnoses and matching treatment strategies. It was the antidote to ‘Complexity’ and the ‘complexity bias’, involving Sat Nav’s for the 10 most commonly occurring disorders. But for 15 years Services have travelled the ‘Complexity’ pathway, with no discernible added value. Unfortunately it does not appear that this bandwagon will stop anytime soon. Perhaps the tide will only turn when people question the economic case for ‘Complexity’?

    Dr Mike Scott

  • Currently, Psychological Wellbeing Practitioners (PWPs) are expected to attend workshops sponsored by the developers of computer assisted CBT (Silver Cloud) and an artificial intelligence company (Limbic). But what will this achieve for the person in the street? At present 1 in 3 people attend just 1 assessment/treatment session in NHS Talking Therapies [Scott (2024)]. Are they seriously more likely to engage more with a computer/AI? The obvious need is for a therapist to give the distressed person the time of day to tell their story and respond appropriately. Reading from a script is insulting. There is no independent evidence for the Service’s claimed 50% recovery rate, the likely true figure, Scott (2018), is a tip of the iceberg recovery rate. To raise these concerns in one of these workshops or indeed in the Service itself, will likely evoke the following response:

    The sales pitch for Workshops is to:

    • boost productivity using AI tools chatbots and digital solutions
    • tailor therapy for neurodivergence with practical, trauma-informed approaches
    • adapt to new key performance indicators
    • determine suitability for low intensity CBT
    • consider that all-comers should get low intensity CBT first.
    • determine unsuitable referrals
    • manage complexity

    PWPs have, in necessary pursuit of CPD, entered a ‘brave new world’. But this is pure marketing, there is no evidence-base that these aspirations can be delivered in the real-world. Rather the PWPs are likely to continue to function in a parallel universe to that inhabited by most clients. Whilst attendees are encouraged to learn from each others ‘best practices’, dissent from NHS Talking Therapies modus operandi is unlikely to be tolerated, the ‘thought police’ will be in operation.

    Overall there is as I said in a recent post a ‘I’m sorry, I don’t have a clue’ response from the gatekeepers to psychological treatment. There is a pressing need to review fundamental methodology.

    Dr Mike Scott

  • Not a Radio gameshow, but NHS Talking Therapies practitioner’s felt responses to encountering clients who mention trauma,  revealed in a paper by Kerr et al (2025) in the The Cognitive Behaviour Therapist. Psychological Wellbeing Practitioners (PWPs) are the gatekeepers into the Service,  but in  this study they report being out of their depth, unable to differentiate trauma responses. With no training in reliable diagnosis or trauma treatment. But clients are unaware of this. Whither transparency?

    The authors of the paper are Oxford University researcher’s, they studiously avoid saying that the assessment process in NHS Talking Therapies is fundamentally flawed. To do so may scupper their access to data from the Service and embarrassingly, underline that the assessment process employed, originated with one of their colleagues.

    I assessed John (some details changed to protect confidentiality) who fell from a ladder, and was unable to work as a painter and decorator for many months. He developed depression and panic disorder. John was routed via his GP to NHS Talking Therapies, where he had 7 sessions of ‘low intensity CBT’. He was deemed by the PWP to have ‘PTSD/like symptoms’ and put on a waiting list for ‘high intensity CBT’ , where he is currently languishing. John hadn’t been informed that he would likely have trauma-focussed CBT involving repeated reliving of his fall! There was no precision in the PWP’s assessment. The fall was undoubtedly a stressor, but it was very unlikely that it was the extreme stressor required for a diagnosis of PTSD. I conducted a standardised diagnostic interview for PTSD, enquiring about each symptom and whether each cleared a threshold for significant impairment. He quite definitely did not have PTSD.

    But most people have experienced a trauma in their life, and those who go to NHS Talking Therapies will be no exception. If the gatekeepers, the PWPs, are at sea on this, the whole service is questionable.

    Dr Mike Scott

  • That is the prediction of Alan Frances in a just published British Journal of Psychiatry paper. He is a leading US psychiatrist and a major figure in the development of the DSM criteria. Yet just returning from the European Conference on CBT at the University of Glasgow, I did not come across any mention of AI. There were attendees from over 60 countries and over 2000 CBT afficionados. But there is clearly a major issue to address in terms of clients and psychological therapists. Bluntly, jobs, could be on the line.

    Alan Frances notes that psychotherapy and guidance on managing everyday difficulties is the most common use of AI. Frances believes that AI is just as good as traditional treatments for mild anxiety and depression. He suggests that an advantage of AI is that it makes engagement of clients easier, in that it is available 24/7 and that there is no shame in disclosing highly personal information to a machine. Frances notes massive investment in the development of AI for mental health problems with an associated likelihood of massive marketing. In this context empirical investigation of outcomes is likely to be given scant attention. Governments and service providers are likely to be carried away by the anticipated huge savings in staffing costs, ignoring misgivings about effectiveness. Frances argues most clients will gradually opt for the accessability, convenience and reduced costs of artificial intelligence. Artificial intelligence chatbots place a premium on pleasing the user, as this is the most likely to be financially lucrative, rather than being governed by what is clinically important. The developers of AI therapy have little accountability.

    But the evidence base for the effectiveness of AI in treating mental health problems is weak. Frances’s contention that AI is as good for treating mild anxiety and depression as traditional treatments is unproven. Nevertheless, my own research Scott (2018) on NHS Talking Therapies clients, suggest only a tip of the iceberg recovery rate. It would be no surprise to find that AI therapy has comparable efficacy, or more accurately lack of efficacy. But AI may be more engaging. For every 2 people having 2 or more treatment sessions in NHS Talking Therapies 1 has just one assessment/treatment session Scott (2024). To my knowledge AI has no demonstrated efficacy for severe mental illness. It could be argued however that in this context it is no different to traditional treatments. 

    In essence AI is like a photocopy of ‘good practice’. But the presumption is that the ‘good practice’ has first been rigorously evaluated. Importantly that the supplied data was ‘falsifiable’. If data from routine psychological therapy was used as the database, with Service providers stipulating what is ‘good practice’, then no steps have been taken to ensure that it was possible to prove that the service was ineffective. In such instances any observed improvement could be due to time, attention and a credible rationale. Thereby making AI a worthless photocopy. In computer terminology It is a matter of GIGO, ‘garbage in and garbage out’. Routine psychological treatment will be, sooner or later in the ‘dock’, whilst there are no signs of it being able to make a ‘robust’ defence, my guess is that the ‘judge’ [political masters, Department of Health] will be swayed by the plausibility of the case presented and the new-found ease of providing services. The acoustics will prevent listening to the voices of clients. There needs to be a ‘wake-up’ call to examine these issues.

    Dr Mike Scott

  • Just looked at the conference program that I will be attending, remotely, later this week. What struck me is that the overwhelming majority of presenters are University-based, with a few non-university-based presenters, running skills classes. But I guess only the tip-of-the iceberg of attendees will be University-based. I am sure that this isn’t a deliberate ploy on the part of Conference Organisers, But it is an issue that has not been recognised since the inception of BABCP, much less addressed.

    Lived experience seems not much in evidence. The danger is that suggested interventions will be eminence-based rather than evidence-based. With University staff often charismatic. The only protection for most attendees is to use the mnemonic PICOT, in gauging the relevance of what is presented.

    Asking the basic question ‘what proportion of people with the said difficulty/disorder, are back to their old selves following this intervention and for how long?’ Presenters may well fudge their response.

    Oftentimes I have found I’ve picked up something really useful at a Conference by attending something that is off my ‘beaten track’. Hopefully this will happen again.

    Years ago there was the CBT Cafe as a CBT discussion group, but this fell into dis-use – I think because of the pre-eminence of some contributors. I was disappointed that it has not been resurrected in some form on the new BABCP website. But I can see that the task of moderation is onerous, nevertheless there is danger for any Organisation if there is not open discussion.

    Dr Mike Scott

  • ‘Therapist responsiveness and tailoring treatment to themselves’. These are the main findings from a qualitative study of 148 NHS clients, who had psychological  therapy,  just published by Li et al (2025).

    Therapist responsiveness means actively listening to the client’s concerns and responding to them. In terms of the social psychology ‘Elaboration Likelihood Model of Persuasion‘, the concerns of the client can be peripherally processed, by the therapist, with effortful processing reserved for a specific identified protocol for a disorder.  In the words of the title of this paper ‘It felt like I was being tailored to the treatment rather than the treatment being tailored to me’. 

    A common scenario is a single Mum living in appalling housing conditions. Her children’s asthma worsened by the mould. The Social landlord has repeatedly failed to address matters. She is severely depressed and the therapist is preoccupied with a Behavioural Activation (BA) programme. Notwithstanding, that she is too embarrassed to invite family/friends to her home.

    Effortful processing on the part of the therapist would have been discussing possible legal action against the Landlord.

    But the job of triaging clients in NHS Talking Therapies, in practice, usually falls to   Psychological Wellbeing Practitioners or a social science graduate, neither of whom are trained therapists.  They do not have the competence to deal with the psychological and social.  The ‘PWPs’ are under pressure, with typically just 6 treatment sessions at their disposal. They may see this single Mum as unsuitable for the service or plough on with a short course of BA. However, the Mum is likely to attend only one assessment/treatment session. With regards to depression for every one person attending one session, two people go on to have 2 or more sessions [Scott (2024)]. There is an unrecognised, massive haemorrhaging of clients at triage. Comparable in scale to booked appointments for an assessment/treatment session that is not attended. 

    There is however a fundamental flaw in the Li et al (2025) paper,  they state that the Improving Access to Psychological Therapies Programme (the predecessor of NHS Talking Therapies) has been ‘internationally recognised for its effectiveness in psychological care on a large scale’, citing David Clark the prime mover in the development of IAPT and Wakefield et al (2021). Neither of these sources can claim independence of the Service, the corresponding author in the Wakefield et al study was an IAPT programme director. I wrote a rejoinder to the Wakefield et al study in the British Journal of Clinical Psychology,  in which my own analysis of 90 IAPT cases as Expert Witness to the Court, showed that only the tip of the iceberg recovered [Scott (2018)].

    Dr Mike Scott

  • Group treatments are a response to long waiting lists for individual therapy. They can also help to normalise a child’s difficulty and help parents to feel less isolated. In recent years there has been a trend away from diagnosis specific groups to transdiagnostic groups. But a just published paper in the Clinical Psychology Review of 80 studies suggests they are no more effective than active controls. Rather than answer whether ‘going transdiagnostic’ represents added value over traditional disorder specific interventions, (the appropriate counterfactual), these authors simply make a case for more rigorous research.! In this vacuum marketing is likely to win – 29 of the studies were by author’s of the developed protocols.

    Caution is needed in translating these studies to routine practice:.

    • Only 1 of the 80 studies was conducted in the Uk, a mentalisation based treatment for youth.
    • Half the studies had parental involvement an average of 8 sessions. The children had on average 11 sessions.
    • 59 of the studies were of CBT consisting of mindfulness, emotional regulation and cognitive restructuring. But the author’s of the studies, self-labelled their interventions. The meaning of ’emotional regulation’ was unclear.
    • 14 of the studies involved children with ASD or ADHD.There is a bottleneck in the UK for ASD and ADHD assessments, as they are the passport to extra help at school. In this context Group Transdiagnostic approaches are likely to be heralded as, at least a ‘first-line approach’. Welcomed by schools to head-off the extra staffing investment this ‘passport’ would confer. Parents may be led to believe that their children are getting a new evidence-based treatment. Service providers are likely to seize upon transdiagnostic group interventions as a simple solution to the pressure of referrals. There is a need to look beyond convenience. 

    In ‘Simply Effective Group CBT’ London: Routledge (2011) I detailed how it was perfectly possible to address the commonplace additional disorders in this modality with Adults, without losing focus on the principal disorder. But over the last decade the conventional wisdom has been to label the additional disorders as constituting ‘complexity’, with a supposed need for a ‘complex’ intervention. However there is no empirical evidence that this shift in focus has resulted in better group CBT for adults. Unfortunately the ‘complexity’ myth has spread to interventions with the young.

    Dr Mike Scott

  • According to a recent British Journal of Psychiatry editorial, the absence of biomarkers for any psychiatric disorder has been used to call into doubt, the whole diagnostic enterprise in mental health. But the editorial suggests that the absence of biomarkers in general medicine is commonplace. Whilst a biomarker can indicate a mechanism of dysfunction, it cannot determine whether pathology exists. The determination of pathology cannot be outsourced to biology.

    The assessor is an actor in the determination of pathology and a conduit for societal values. It appears that in the very act of trying to gauge pathology, what is the focus is being changed. The situation seems akin to in physics, that the more that one determines the position of an atomic particle the more uncertainty there is as to its’ velocity and vice versa (the Heisenberg Uncertainty Principle). It appears that for the foreseeable future, it’s likely to be a matter of clinicians muddling along with diagnostic categories, that may not be carved into nature, in the absence of a better way of categorising difficulties and corresponding treatment options. A great deal of humility is called for, not the strong suit of many psychiatrists or psychologists. 

    NHS Talking Therapies operates a free for all, its’ clinicians ascribe a diagnosis based on ‘ICD-codes’ and not on any standardised diagnostic interview. Making for highly questionable reliability. Nonetheless luminaries of ‘critical psychology’ have used NHS Talking Therapies data to substantiate the very significant withdrawal effects from antidepressants.Whilst I don’t doubt that this is a very real issue, they don’t at all raise the issue of the reliability of NHS Talking Therapies data. One is reminded of certain President’s inability to criticise other Presidents.

    Dr Mike Scott

  • I-Phone now provide the facility for users to complete the PHQ-9 and GAD-7 and get feedback on the results https://support.apple.com/en-gb/guide/iphone/ipha8d27408f/ios. They caution that the results only indicate the ‘risk’ of developing anxiety (measured by GAD-7 score if over 13) and depression (measured by PHQ-9 score if over 18). But they provide no quantification of the ‘risk’. The I-phone suggests that those with PHQ-9 scores of 10 or more or GAD-7 scores of 8 or more should consult their GP/Specialist. If only a small proportion of the 1.5 billion I-phone users did this, it would likely overwhelm front-line services, crowding out those who do really need a clinician’s help.

    Most people appreciate that every drug has a side effect. But what is not appreciated is that a psychometric test, administered in isolation, has serious adverse consequences – false positives, for every person correctly identified between 1-4 other people are mislabelled and likely mistreated. I explain how this comes about at the. end of this blog. Suffice it to say for now that there is no evidence that scores on these tests, administered without specification of context, predict anything.  Scores can be high for myriad other reasons, than depression or anxiety, e,g exam results, strained relationship, housing disrepair, disorders other than anxiety or depression. For example, a just published study on bulimia in Japan, found a quarter were suffering from depression. The public are likely to dismiss the smallprint about ‘risk’ because of a lack of understanding, unwittingly creating a ‘mental hypochondria’. But don’t worry, before long Apple will develop an App for measuring this construct, despite it lacking any validity. The results can then be easily integrated into AI generated therapy. It is no wonder that Illinois has already banned AI Therapy. 

    Tests are very good on sensitivity, i.e they identify the vast majority ( about 90%) of people who have a disorder. The downside is that they have poor specificity (about 60%),correctly identifying barely half of those who do no t have the disorder. Imagine a 100 people are assessed with a ‘gold-standard’ diagnostic interview, it will likely identify 10 of them as being depressed (top left hand column below) and 90 as not having depression (top right hand column below).Those who do not have a disorder are far more numerous those who have a disorder. Applying the test to column one, 9 out of 10 would be above the cut-off and 1 below. Moving now to the 2nd column, the 90 people who do not have the disorder, the test only gets 60% of these right, ie 54 people, and gets 36 people wrong. Thus there are 9 people that the test got right versus 36 people that it got wrong, thus for every person it got right, it got 4 wrong.

    Imaginary Sample of 100 clients

    Depressed n=10Not depressed n=90
    Above the cut-off n=9Above the cut-off n=36
    Below the cut-off n=1Below the cut-off n=54

    The adverse consequences of the test are that approximately 4 times as many people are screened positive as have the disorder. Thus by itself the test would lead to rampant invasive treatment. Using the analogy of breast screening many more tests would be conducted before surgery or chemotherapy. However with the sole test used in NHS Talking Therapies there are no such checks and balances.

    When the PHQ-9 and GAD-7 have been evaluated alongside a diagnostic interview, the tests give a prevalence rate two to three times higher. Patients with principal diagnoses of post-traumatic stress disorder, obsessive compulsive disorder, specific phobias and adjustment disorders commonly score above thresholds on these tests. So that when these tests are reified the test completer Is nudged in the wrong treatment direction,.

    Completing these tests fits neatly into the time slot of the typical GP consultation time of 10-15 minutes. Psychological therapists will likely not see the use of the I-phone in the therapeutic hour as too burdensome, conferring a ‘scientific glow’ to proceedings, even if they do not know what they are measuring. 

    Using the PHQ-9 the natural recovery rate for depression amongst GP patients administered an antidepressant was 47% within 3 months [Moore et al (2012)]. So that ‘watchful waiting’ is probably best in the 1st instance , waiting for the Storm to pass rather than heading into the eye.

    Dr Mike Scott