The Near Extinction of CBT

Evidence-based psychological therapies are near extinction. Their demise began in 2008 with the inception of the Improving Access to Psychological Therapies (IAPT) service. Aided and abetted by the British Psychological Society’s validation of IAPT’s Psychological Well-being Practitioner’s (PWPs) training programmes and the service’s fellow traveller, the British Association for Behavioural and Cognitive Psychotherapy (BABCP). Gone is the welcoming open door and the careful distillation of what ails the client, instead there is a 30 minute+ telephone conversation, with a third of people then not going beyond one treatment appointment.


The public most commonly receive PWP ministrations when they seek NHS psychological help. But the PWP’s do not follow any treatment protocol for any disorder, indeed they do not make diagnoses. How then can they be said to deliver CBT? By the spurious claim  that they can select a CBT strategy which is sufficiently potent. But they furnish no evidence of systematically following any strategy, notwithstanding that there is no evidence that CBT strategies delivered as stand alone interventions make any real world difference. The PWP’s deliver the Alice in Wonderland, Dodo verdict on CBT strategies ‘all are equal and must have prizes’. Raising the question ‘is CBT as dead as the Dodo?’



Where else might CBT be found? It is not impossible for it to be delivered in IAPT’s high intensity service, but few of its practitioners conduct a reliable standardised diagnostic interview which is the foundation for delivering CBT.  The  treatment integrity of high intensity CBT interventions has never been assessed.  No steps have ever been taken to ensure clinicians are dovetailing diagnosis appropriate treatment targets with matching treatment strategies. Is CBT to be found in private practice? It is possible, but private organisations have largely sought to ape IAPT in the mistaken belief that this confers credibility. Are the chances of finding CBT in private practice comparable to finding life on Mars?

Is CBT alive and kicking in secondary care? Here we enter the muddy waters of clients who might traditionally be regarded as having personality disorders (PD). But there is an understandable reluctance to use the term PD because of the associated stigma and because historically use of such a term has consigned people to the dustbin. Nevertheless Sperry and Sperry (2016) have produced the 3rd Edition of CBT for DSM-5 Personality Disorders (Routledge) but it is eminence-based rather than evidence-based. It is light on outcome studies. I struggled to find any where there was independent assessment of outcome by blind raters, use of an outcome measure that clients would regard as a minimally important difference and evaluations by those other than the creators of the protocols. It is a free for all with strategies such as ‘thought stopping’ recommended, without specification of any contraindications such as PTSD or OCD. Only eclipsed by recommending solution focussed therapy for anxiety. If clinicians in secondary care operate on this text it is very different to Beck’s own work on CBT for personality disorders. But no typology of what clinicians say they do and what they actually do in secondary care has been produced. Tertiary care seems preoccupied with crisis management and is not guided by any recognisable CBT protocol.

In neither primary or secondary care is there a differentiation of treatments or clients. Thus in the UK it is impossible to answer the question of ‘What Works With Whom?’. This leaves clinicians up a creek without a paddle.

Dinosaurs may have been wiped out by an asteroid hitting the earth 66 million years ago, but life survived, doubtless CBT will survive the impact of IAPT, but it is a close call and it is likely going to be down to individual practitioners doing what they know to be best for their clients.


Dr Mike Scott

Low Intensity CBT Works ‘Just like That’

as the magician Tommy Cooper would have said. The Improving Access to Psychological Therapies Service (IAPT) invites its’ paymasters Clinical Commissioning Groups (CCGs)/NHS England to fund a simple and cheap solution to mental health problems. If it sounds too good to be true, it probably is. But this hasn’t stopped IAPT becoming the over a £1bn a year magnet for investment and all without independent assessment.

Here is an extract from a Psychological Wellbeing Practitioner’s (PWP) letter to a GP:

Stress Control Course undertaken outcome was successful evidenced by the first and last questionnaire




Nov 2018



Jan 2019



The PWP is unaware that a score on a psychometric test is not an evidence-based construct. Such scores are not specific to anything. They cannot be used as a surrogate for a diagnosis.  ‘Stress’ is a fuzzy, the terms usage in this context, resembles Alice in Wonderland where words mean whatever you want them to mean.  Further, a change in the score is not evidence that the person’s needs have been met.

The mechanism’s of action in the original randomised controlled trials of CBT for depression and the anxiety disorders were clearly stipulated. As were outcomes e.g loss of diagnostic status as assessed by independent clinicians. But in the low intensity interventions there is no specification of an evidence-based mechanism for change. In effect we are invited to believe in the magic, it works just like described in the letter abstract. One can only gasp at the incredulity of CCGs and wonder what agenda they are working on. But the British Association for Behavioural and Cognitive Psychotherapy (BABCP), the ‘Lead’ organisation for CBT practitioners has a special section for PWPs and IAPT rejoices that its’ low intensity CBT courses are validated by the British Psychological Society. They have failed clients abysmally. 

Dr Mike Scott

Another Nail In The Coffin of IAPT

A year ago the British Journal of Clinical Psychology published my paper ‘Ensuring that the Improving Access to Psychological Therapies (IAPT) programme does what it says on the tin’  60(1), 38. This month in the Journal there is a further damning indictment by Martin et al (2022) ‘Improving Access to Psychological Therapies (IAPT) has potential but is not sufficient: How can it better meet the range of primary care mental health needs?’ 61, 157–174, DOI:10.1111/bjc.12314.

Here are the main points from Martin et als’ BJCP paper:

  •  Improving Access to Psychological Therapies(IAPT)has significantly increased access to psychological therapies within primary care over the last decade, though it is unclear whether its interventions are sufficiently tailored to meet the actual levels of complexity of its clientele and prevent them from needing onward referral to secondary care as originally envisaged.
  •   Given the ongoing focus on and investment in IAPT informed developments into long-term conditions and serious mental illness, this review considers whether additional elucidation of the model’s original objectives is required, as a precursor to its expansion into other clinical areas.

  •   There view indicates that there is a stark lack of data pertaining to the generalisable, real-world clinical benefits of the IAPT programme as it currently stands.

  •   Recommendations are provided for future areas of research, and practice enhancements to ensure the value of IAPT services to clients in the wider context of NHS mental health services, including the interface with secondary care, are considered.


The British Association of Behavioural and Cognitive Psychotherapies (BABCP) ought to look seriously at the promotion of its’ IAPT comic ‘CBT Today’. Interestingly in its’ recent issue it managed to omit that I was one of those who made a submission re: the proposed NICE Guidance on depression. Further, only one of the others who made submissions were given their adjectival title, the leading light in IAPT. The British Psychological Society (BPS) should reconsider its validation of low intensity IAPT courses, in the absence of any credible evidence base on real-world effectiveness.

Dr Mike Scott

The Extraordinary Claims and Behaviours of IAPT’s Backers

who exhibit power without accountability. They include companies such as SilverCloud and limbic and professional bodies such as the British Association for Behavioural and Cognitive Psychotherapy (BABCP) and the British Psychological Society (BPS).

I reported SilverCloud to the ASA over its claim of ‘up to 70% clinical recovery rates’ for its computerised CBT. Unfortunately they could not act on it as it comes under Irish jurisdiction. The matter has  been passed to the Irish ASA, from whom I have heard nothing. The reach of SilverCloud is extensive, with its’ claim to be “supporting 80% of the NHS Improving Access to Psychological Therapies(IAPT) services”. It is a major financial backer of IAPT workshops. But there has been no independent verification of SilverCloud’s claimed recovery rates.

IAPT workshops are also now funded by limbic ‘An A. I. assistant for clinical assessment in IAPT – improving access, reducing costs and freeing up staff time’. Recently the British Psychological Society Journal the Psychologist devoted an article to the claims of the CE0 of limbic. I protested, and furnished a critique which the Editor declined. I note that in the current issue of the American Journal of Psychiatry that there is a paper by IAPT researchers Delgadillo et al 2022 JAMA Psychiatry. 2022;79(2):101-108. doi:10.1001/jamapsychiatry.2021.3539 published online December 8, 2021 in which they have been unable to substantiate the claims of limbic.

But Delgadillo et al (2022) do claim a 7% increase in the likelihood of recovery if IAPT therapist use the limbic algorithm i.e inputting data on depression, anxiety, history etc to determine whether the particular clients needs are better met by IAPT standard stepped care or by a stratified procedure where clients are allegedly better matched to high or low intensity CBT initially. However they do observe that the apparent difference could be due the therapists involved in stratification devoting more time to clients!

Delgadillo et al (2022) accept without question IAPT’s definition of recovery, a change of score on a self-report measure, the PHQ9, to below caseness. They fail to point out that their metric does not a) involve independent assessors to counter the demand characteristics involved in usage of a self-report measure i.e the focus on this measure in client-therapist interactions b) the IAPT data provides no indication that clients see the claimed changes as clinically meaningful, i.e back to old self or best functioning c) symptoms of depression and anxiety wax and wane, so that any improvement on a self-report measure can be simply a flash in the pan, particularly when people present initially at their worst. It has to be determined that any change is lasting e.g at least 8 weeks. It appears that Delgadillo et al (2022) simply rejoice in the large data set furnished by IAPT, it is a case of ‘never mind the quality, feel the width’.

When the power holders collude in this way, it is difficult make headway. I think limbic should also be reported to the ASA and BABCP and BPS should be asked to justify their commitment to Psychological Wellbeing Practitioners (PWPs), the deliverers of low intensity CBT – it looks suspiciously like cronyism, however unintentional.

Dr Mike Scott

People Cannot Benefit from a Treatment To Which They Have Not Been Exposed – The Undermining of IAPT

The Improving Access to Psychological Therapies (IAPT) Service does not assess treatment fidelity. Thus, there can be no certainty that clients receive an evidence-based treatment treatment.  IAPT therapies are not EBTs. Despite this, the major funder of IAPT training days SilverCloud, claims on its’ website ‘up to a 70% real-world recovery’ using its computer assisted products, for all common disorders except PTSD and OCD!  The Advertising Standards Authority need to look at this, the ASA has a complaints form that can be completed online. SilverCloud’s UK address is Suite 1350, Kemp House, 152 City Road, London, EC1V 2NX., My own study of 90 IAPT cases suggests just a 10% recovery rate, Scott (2018)

IAPT 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. see SilverClouds Space for Depression programme   NICE Guidance ‘Space from depression for treating adults with depression’ Medtech innovation briefing published May 7th 2020. Strangely the NICE IAPT Expert Panel concluded that the case for adoption is ‘partially supported’ despite in the body of report noting lower depression scores, at the end of treatment for the clients of therapists who did not use the computer assisted CBT. An example of spin and conflict of interest.


The SiverCloud website cites 10 references appearing in peer-reviewed journals to support its work.  But none of the studies cited by SilverCloud involve blind independent assessors of outcome using a ‘gold-standard’ diagnostic interview. In the cited review study by Wright et al (2019) Wright JH, Owen JJ, Richards D, et al. Computer-assisted cognitive-behavior therapy for depression: a systematic review and meta-analysis. J Clin Psychiatry. 2019;80(2):18r12188 the third author is employed by SilverCloud.

 ‘Real-world’ recovery represents a change that a client would care about, such as no longer suffering from the disorder that they were suffering from before treatment or a return to best functioning. In a footnote SilverCloud defines recovery as ‘Moving from clinical caseness to non-caseness, i.e. lowering the score on PHQ-9 and GAD-7 from above the clinical threshold to below the threshold’. Such changes are meaningless to clients they are not ‘real-world’.

Here is what one client told  me:

‘I found Silvercloud ineffective, generic and not tailored to my personal situation. It wasn’t engaging or helpful and as such I didn’t engage with the website very much. Consequently, the following weekly call with the IAPT therapist  were sometimes made difficult by the fact I hadn’t completed the same questionnaire as the week before or read through articles. I wanted to talk about my situation, my feelings and find out why I was feeling the way I was, but I felt I was just being led back to using the online SilverCloud resource.

‘It was in 2017 that my doctor suggested I try SilverCloud online CBT with telephone support and in September 2017, I started speaking to another IAPT counsellor. He seemed to be a very nice man. After a few weekly calls, he stated that he didn’t believe I was depressed and so he changed the original Silvercloud course I had started and reset it back to a new series of 6 sessions. The weekly calls lasted between 20 minutes to an hour depending on what we discussed, but always concluded with him asking me to log onto SilverCloud and work my way through the programme before our next call. After the requisite 6 sessions finished in February 2018, that was it! No answers, no tools to help me cope, just signed off, discharged, but told I had 12 month access to SilverCloud. I haven’t used the resource since’.

In general the claims of clinicians and supervisors with regards to treatment fidelity do not match those of independent blind-raters [ Waltman et al (2017)], there are vested interests at play.

The author knows of no study of low intensity CBT (guided self-help, group psychoeducation, computer assisted CBT) that has assessed treatment fidelity. Usage of a manual does not guarantee treatment fidelity. Approx. three quarters of IAPT clients receive low intensity intervention on entry to the Service [Davis et al (2020)].].

IAPT’s approach ostensibly depends on the results of randomised controlled trials of CBT, but a study of remission rates in CBT for anxiety disorders (including OCD and PTSD) Levy, Bryan and Tolin (2021) showed that in half the studies (8 out of 17) there was a high risk of bias because of a failure to address treatment fidelity. Further in 7 of the 17 studies there was a high risk of bias because of the failure to use blind assessors. [A re-view of psychotherapy trial reports published in 6 top psychiatry journals in 2017 and 2018 revealed that only 59% of the included trials reported adequate blinding of outcome assessors Mataix-Cols et al (2021)].].Thus, the research base that IAPT draws upon is far from rock solid.  The remission rate in rcts for anxiety disorders is approx. 50% [ Springer et al (2018)]and this is the ‘gold standard’. But IAPT claims comparable results despite a total disregard for blinding and treatment fidelity! The faked goods ought perhaps to be reported to Trading Standards as well as ASA, in lieu of any interest in the matter from the British Psychological Society (BPS) or the British Association for Behavioural and Cognitive Psychotherapy (BABCP)!

The real story of SilverCloud is that it provides morsels of CBT when what is really needed is a proper meal. It is insulting to clients to in effect say ‘let’s see how you get on with morsels and then we will see about a proper meal’.


Dr Mike Scott

It’s A Myth That The Improving Access to Psychological Therapies (IAPT) Service Pays for Itself

IAPT has flourished over the last decade by proclaiming that it pays for itself [see Layard and Clark’s book Thrive  (2014)]. It has been music to the ears of politicians, NHS England and Clinical Commissioning Groups  but none, including the National Audit Office, has bothered to question it. Despite the £1bn price tag this year, see footnote 1. Anyone with the temerity to raise doubts, risks being accused of lacking a commitment to mental health, a pre-requisite of being considered progressive, whatever one’s political hue. 


When will the funding and professional bodies such as the British Psychological Society (BPS) and British Association for Behavioural and Cognitive Psychotherapy (BABCP) see that the ‘Emperor Has No Clothes’? IAPT claims the service pays for itself by getting people off unemployment benefit (16.8% of IAPT clients) Davis et al (2020)  and/or long term sick or disabled benefit (6.9% of IAPT clients).   It is therefore a change in the employment status of minority of IAPT clients that may justify the belief that the service pays for itself. But further elaboration of this population shows that the proportion of clients who could make an economic difference is smaller still. Further when the psychological mechanism by which a change of occupational status may operate is considered, it is improbable that the service pays for itself.  

 IAPT could in principle get 20-25% of clients off benefits. Assuming the target clientele this year is 20%, i.e 0.3 million people, how would the service pay for itself?  Well 40% of IAPT clients do not attend their 1st treatment appointment, so only 0.18 million will be exposed to an IAPT treatment therapist. Of these 42% attend just one treatment appointment, thus 0.1044 million have exposure  to IAPTs treatments and are in the categories of unemployed or long term sick, and potentially might have their employment status changed by the Service i.e 104,440. Those undergoing IAPT treatment ( defined by the Service as attending 2 or more treatment sessions) have an average of 8 treatment sessions in 2018-2019 Saunders et al (2020) but the unemployed and those on long term sickness benefit are less likely to attend a treatment session, Davis et al (2020), as are those who have been referred previously. Thus one might expect this 104,440 to attend a mean of 6 sessions and treatment typically spans 12 weeks according to Saunders et al (2020) . But the population who may return to employment is smaller still because of the following considerations:

  1. There will be a sub-population of the ‘unemployed’ whose unemployment is  related to a work related negative life event, e.g now being physically unable to do the manual work they were employed to do or maltreatment at work. It is difficult to see how 6 sessions of psychological therapy  delivered over 12 weeks would change the diagnostic status of this sub population. There is absence of evidence that such a dosage of psychological therapy can change the employment status of this sub-population. If the sub-population of clients for whom work has been an iatrogenic factor in their debility, are excluded from the analysis, then the population that IAPT’s ministrations could conceivably address is much less than 100,000.
  2.  There will be a further ‘sub-population’ of the unemployed for whom work within their training is simply not available e.g a redundant fisherman. IAPT does not have the resources to conjure up new opportunities, albeit it might direct a client towards re-training.  

Thus the range of action of IAPT with regards to employment status is very limited and even more so when one considers by what mechanism could the typical 6 sessions change employment status over the 12 week span? To return a person to occupational functioning means addressing three key areas a) persistence – the ability to persist with a task b) pace – the ability to complete a task in a timely manner and c) adaptation – the ability to handle the inevitable hassles of the workplace. There is no evidence that IAPT specifically targets these difficulties or has provided training in tackling them. Neither has it been demonstrated that 6 sessions of psychological therapy can resolve such difficulties in 12 weeks and even less evidence as to whether such treatment is enduring.

IAPT lacks the potency to make a real world difference to the unemployed and those on long term sick. Layard and Clark (2014) muddy the distinction between the power of evidence-based psychological therapies and the power of their offspring, IAPT. It can be objected that IAPT pays for itself by increasing the productivity of those already employed, rather than by changing occupational status. But there is no evidence that it does so anymore than the pre-IAPT counselling services.

IAPT’s claim that it changes the employment status of its’ clients is akin to a Dickensian Government’s claim that Workhouses resolve employment issues.

Footnote and reference


  1. According to The IAPT Manual 2021 the target for 2021 is 1.5 million clients at a cost of £680 per client [data from Clark (2018)] making the anticipated cost of the service this year, £1.02 billion.
  2. Layard, R and Clark, D.M ( 2014) Thrive: The Power of Evidenced-Based Psychological Therapies Penguin Limited

Dr Mike Scott


Notice Served On IAPT’s Claim

of a 50% recovery rate. The Editors of Lancet Psychiatry S2215-0366(21)00123-1 have challenged researchers to demonstrate that an acclaimed intervention makes a difference that service users would recognise. Thus making the consumer of mental health services centre stage rather than a change in score on a test. In addition researchers are asked to justify their primary outcome measure. In interpreting test results the Editors insist that  author’s must clarify what a change of X would mean to a service user as opposed to a change of Y. A recently published paper in the Journal, using IAPT data, S2215-0366(21)00083-3 would probably not have been published, if it had not been accepted just before the new guidance was implemented. If other Journal editors follow suit, IAPT’s wings will have been clipped over the claims of IAPT and its’ fellow travellers, such as the British Psychological Society (BPS) and the British Association of Behavioural and Cognitive Therapies (BABCP).  There has been a dereliction of duty by BPS and BABCP.


In this connection I have had the following correspondence with the Lancet Psychiatry  Editors:

My letter

When A Difference Makes No Difference

In June this year the Lancet published guidance [Boyce et al (2021)] for mental health researchers to ensure that the primary outcome measure employed in a study needs to be meaningful. Researchers were asked to a) justify their choice of an outcome measure and b) specify what a change of X or Y on a measure would mean for a service user. Contemporaneously, Lancet Psychiatry published a study by Barkham et al (2021) that made no attempt to address the Editor’s expressed concerns.

Barkham et al (2021) chose to adopt the Improving Access to Psychological Therapies (IAPT) primary outcome measures the PHQ-9 [Kroenke et al (2001)] and GAD-7 [Spitzer et al (2006)], without any discussion. There is no comment that these are self-report measures, subject to demand characteristics and that changes are impossible to interpret without comparison to an active placebo treatment.

The Barkham et al (2021) study involved comparison of person-centred counselling and cognitive behaviour therapy (cbt) in a high intensity therapy service delivered by IAPT. Curiously patients were screened for the study using the Clinician Interview Schedule Revised but neither this nor any standardised diagnostic interview was used as an outcome measure. Why such apparent blindness? The answer is apparent reading the declaration of conflicts of interest, the authors are either devotees of person-centred counselling or have links with IAPT. Their take home message is that person centred counselling might be better than CBT for depressed patients. But there is no attempt to address the question of what proportion of patients lost their diagnosis status and for how long, as determined by an independent blind clinical assessment using a standardised interview. Service-users interests are ill-served by this type of study which additionally ignored data that suggest the recovery rate in IAPT is just 10% [Scott (2018)].


Barkham, M., Saxon, D., Hardy, G. E., Bradburn, M., Galloway, D., Wickramasekera, N., Keetharuth, A. D., Bower, P., King, M., Elliott, R., Gabriel, L., Kellett, S., Shaw, S., Wilkinson, T., Connell, J., Harrison, P., Ardern, K., Bishop-Edwards, L., Ashley, K., Ohlsen, S., … Brazier, J. E. (2021). Person-centred experiential therapy versus cognitive behavioural therapy delivered in the English Improving Access to Psychological Therapies service for the treatment of moderate or severe depression (PRaCTICED): a pragmatic, randomised, non-inferiority trial. The lancet. Psychiatry, 8(6), 487–499.

Boyce, N., Graham, D., & Marsh, J. (2021). Choice of outcome measures in mental health research. The lancet. Psychiatry, 8(6), 455.

Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a brief depression severity measure. J Gen Intern Med 2001;16: 606–13.

Scott M. J. (2018). Improving Access to Psychological Therapies (IAPT) – The Need for Radical Reform. Journal of health psychology, 23(9), 1136–1147.

Spitzer RL, Kroenke K, Williams JB, Löwe B. A brief measure for assessing generalized anxiety disorder: the GAD-7. Arch Intern Med 2006; 166: 1092–97.


August 13th 2021


Thank you for your letter to The Lancet Psychiatry. We are pleased to see that our initiative re primary outcome reporting has been noticed.  We are applying this now but did not apply it retrospectively to papers accepted before publication. The Barkham et al paper was published online on 14 May, six weeks after the Comment, but was accepted and edited before our new policy was in place. 

For Correspondence, our information for authors states: Letters written in response to previous content published in The Lancet Psychiatry must reach us within 4 weeks of publication of the original item.  We do extend this to after the original item has been published in an issue but I’m afraid that your letter is still outside the window for the Barkham et al paper, so we have decided not to publish it.

Although this decision has not been a positive one, I thank you for your interest in the journal.

Yours sincerely,

Joan Marsh

Joan Marsh MA PhD

Deputy Editor


Dr Mike Scott


On What Basis Are Talking Therapies Out Of Bounds To The Care Quality Commission?

The Care Quality Commission (CQC) has just called the Government to task for blanket Do Not Resuscitates (DNRs) applied at the start of the pandemic. But the CQC is not allowed to investigate the quality of the Improving Access to Psychological Therapies (IAPT) services for those with mental health difficulties. Could there be a more glaring example of the disparity between physical and mental health services?


The IAPT service has had a decade of going under the radar of independent public scrutiny, despite Government expenditure of over £4billion. Strangely the National Audit Office (NAO) has no intention of mounting an audit (see recent post), citing preoccupation with Covid and its’ earlier preoccupations with the collapse of Carrillion and the provision of generic medicines. IAPT is responsible to NHS England but staff at the Department of Health also have key positions in IAPT. NHS England are likely to claim that they are ‘too busy’ to address trivial matters like conflicts of interest, reacting like the NAO. The Government will likewise claim preoccupation to avoid addressing sensitive matters.

The CQC can investigate whether the needs of those in Care Homes are being served and can champion the plight of residents, who is to champion the needs of those with mental health difficulties. Organisations such as Mind often have funding arrangements with IAPT. The  British Association for Behavioural and Cognitive Therapies (BABCP) and the British Psychological Society (BPS) regularly give pride of place to IAPT luminaries with rare opportunities for opposing views to be expressed.  The result is a groupthink within these organisations. 

Dr Mike Scott


The Department of Health Has Failed To Regulate Routine Mental Health Services

Improving Access to Psychological Therapies (IAPT) services are out of bounds to Care Quality Commission inspection.  In 2016 the National Audit Office (NAO) asked the Department of Health to address this issue and it has done nothing.  The Department sets the agenda and budget for NHS England, who in turn do the same with Clinical Commissioning Groups to determine local provision of services. But NHS England staff are lead players amongst service providers, these conflicts of interest exacerbate the parlous governance of IAPT. There is a need for Parliament to step in and take the Department of Health to task.  


Whilst no one doubts the importance of improving access to psychological therapies, it was remiss of the NAO in 2016 to take at face value IAPT’s claim that it had the appropriate monitoring measures in place.  Incredulously the NAO accepted at face value IAPT’s claim that it was achieving a 45% recovery. It is always tempting to look only as far as evidence that confirms your belief. But it is equally important to consider what type of evidence would disconfirm your belief. The NAO has failed to explain why it has not insisted on independent scrutiny of IAPT’s claims. 

The The Improving Access to Psychological Therapies (IAPT) programme has exercised a confirmatory bias in its’ audit by focussing only on self-report responses on  psychometric tests (the PHQ9 and GAD7). The service has never looked at a categorical end point, such as whether a person lost their diagnostic status as assessed by an independent evaluator using a standardised diagnostic interview.

Organisations, are inherently likely to be self-promoting and will have a particular penchant for operating, not necessarily wholly consciously, with a confirmatory bias. It is for other stakeholders, NHS England, Clinical Commissioning groups, MPs, the media, Charities and professional bodies (BABCP  and BPS) to hold IAPT to account. For the past decade they have all conspicuosly failed to do so. How have IAPT evaded critical scrutiny, despite the taxpayer having paid £4billion for its’ services? Friends in high places is the most likely answer. I have called for an independent public inquiry for years and will continue to do so  but there is likely to be an echo of a deafening silence as the only beneficiary would be the client with mental health problems.  

Dr Mike Scott

Mental Health Sufferers Vote With Their Feet and Government Does Nothing At All

 of those who undergo an initial assessment with the Improving Access to Psychological Therapies (IAPT) Service 40% do not go on to have treatment, and about the same proportion (42%) attend only one treatment session, according to a just published study by Davis A, Smith T, Talbot J, et al. Evid Based Ment Health 2020;23:8–14. doi:10.1136/ebmental-2019-300133. These findings echo a study  published last year by Moller et al (2019), on a smaller sample, which suggested that 29% were non-starters and that the same proportion attended only one treatment session. Further scrutiny of the data reveals that about 3 out of 4 people drop out of treatment once begun. Unsurprisingly the authors’s independent study, of 90 IAPT clients, Scott (2018) revealed that only the tip of the iceberg (9.2%) recovered                 DOI: 10.1177/1359105318755264, raising serious questions about why the Government has spent over £4 billion on the service.

What Has Gone Wrong?

Kline et al (2020) consider that at an assessment by a clinician is supposed to: a) provide a credible rationale for the proposed treatment b)  detail the efficacy of the envisaged treatment and c) ensure that the clients preferences are acknowledged. IAPT’ assessments fail on all counts, taking these in turn:

a. If the problem is ill-defined e.g low mood/stress it is not clear what rationale should be presented. It is doubtful that a 30-45 minute telephone conversation can provide sufficient space to define the primary problem and other problems/disorders that may complicate treatment. Initial assessments  of patients for randomised controlled trials of psychological interventions are typically 90 mins plus, if this is the time deemed necessary to reliably diagnose a patient by a highly trained clinician, how can a much less trained PWP do it in less than half the time? Under time pressure a PWP may consider providing a credible rationale is part of treatment not assessment and in such circumstances it becomes more likely that a client will default. 

b. How often do PWPs present clients with evidence on the efficacy of an intervention? Take for example, computer assisted CBT, does the therapist tell the client that only 7 out of 48 of NHS recommended e-therapies have been subjected to randomised controlled trials, ( see Simmonds-buckley et al J Med Internet Res 2020;22(10):e17049) doi: 10.2196/170490 and even in these a gold standard semi-structured diagnostic interview conducted by a blind assessor was not use to determine diagnostic status post treatment, i.e there was no determination of the proportion of clients who were back to their old self after treatment and for how long. Further the e-therapies had average dropout rates of 31%.  They are not evidence based treatments in the way the NICE recommended high intensity treatments are. But approximately three quarters Of IAPT interventions (73%) are low intensity first, with 4% stepped up to high intensity and 20% in total receiving a high intensity intervention Davis A, Smith T, Talbot J, et al. Evid Based Ment Health 2020;23:8–14. doi:10.1136/ebmental-2019-300133

c. Client’s preferences are a predictor of engagement in treatment, but how often is a client given a choice between a low intensity intervention and a high intensity intervention. If both options are juxtaposed choice is likely skewed by informing the client that the high intensity intervention has a much longer waiting time.

Defining A Dropout

The generally accepted definition of a dropout is attending less than 7 sessions [see Kline et al (2020)], it is held that clients attending below this number will have had a sub-therapeutic dose of treatment and are therefore unlikely to respond]. Applying this metric to IAPT’s dataset is difficult as they only report data for those who complete 2 or more sessions and for which the average number of sessions attended is 6, thus the likely dropout rate from IAPT treatment, as most would understand the term, is about 75%.  But IAPT has developed its’ own definition of a completer as one who attends 2 or more sessions. This strange definition serves only to muddy the waters on its haemorrhaging of clients. It makes no sense to continue to fund IAPT without an independent  government inquiry into its’ modus operandi.


An Alternative Way Forward

Such has been the marketing power of IAPT over the last decade, that professional organisations such as the British Association for Behavioural and Cognitive Psychotherapies (BABCP) and the British Psychological Society (BPS) have sat mesmerised, as the Services fellow travellers have dominated accreditation and training.     In ‘Simply Effective Cognitive Behaviour Therapy’  published in (2009) by Routledge, I detailed a very different way of delivering services, that represents a faithful translation of the CBT treatments delivered in the randomised controlled trials (rcts) for depression and the anxiety disorders. Unfortunately it is IAPT’s fundamentalist translation of the rcts that has held sway and has brooked no debate either in journals or at Conferences.


Dr Mike Scott