Government Gives £38 million Extra To A Mental Health Service That Has Already Crashed With Real Life Clients

the typical client presenting at the Improving Access to Psychological Therapies (IAPT) service, 58%, are suffering from three or more disorders , with 14% suffering from two disorders [Hepgul et al 2016]. The interventions delivered by IAPT are most commonly minimalist, little contact with a therapist other than to supervise the client’s use of computerised CBT. Strategies on offer in IAPTs low intensity intervention can be found in any CBT self-help book. It can reasonably anticipated that its’ treatments are destined for failure. 

The complexity of IAPT’s target population is highlighted further by a) the study of Zimmerman et al (2005) which suggests that a third of IAPT clients likely have a personality disorder b) Hepgul et al’s (2016) finding that 21% reported moderate or severe sexual abuse and 16% moderate or severe physical abuse and c) Hepgul et al’s (2016) finding that 18% had alcohol dependence and 7% substance dependence. The IAPT manual (2019) claims  that its’ therapists provide treatment according National Institute for Health and Care Excellence (NICE) guidelines, but the latter recommend treatments for single disorders.  Further IAPT provides no evidence of fidelity to any particular protocol. The NICE recommended treatments are based on randomised controlled trials (rcts) of CBT for single depression and anxiety disorders with limited concessions to  comorbidity, not for the range of complexity encountered in routine practice. Thus whilst it is the the case that on average 50% of those in the rcts for depression and the anxiety disorders are in remission at the end of treatment, it is most unlikely that this could be achieved with the IAPT population, yet the service claims exactly this success rate!  

New IAPT staff are like army recruits in the 1st World War, full of optimism, ill prepared for what they find at the front e.g one in four with alcohol/substance dependence. The IAPT practitioners at the front line have little understanding of the prevalences, multiplicity and complexity  of disorders they will find. Training is insufficient to allow them  to distinguish PTSD from a specific phobia, situational bound panic attacks from panic disorder. This is legitimated by the IAPT Manual that claims the Organisation doesn’t make diagnosis, making treatment like playing roulette. No wonder IAPT therapists are burnt out. 

Unsurprisingly I found that when I examined 90 IAPT clients independently using a standardised semi-structured interview the recovery rate was just 10% Scott (2018) . The Government has never independently audited any of the services that it is spending £500 million on. The Minister for Mental Health, Nadine Dories  has said £2.5 million is to be spent on ‘new approaches to support children who have experienced complex trauma’ she appears not to know that the whole notion of ‘complex trauma’ is a matter of considerable debate. Doubtless many have been stressed by the pandemic, but it would make more sense to wait for the storm to pass and assess the storm damage before throwing money at it. There is a danger in pathologising normal responses to an abnormal  situation and not being guided by evidence based interventions.

Dr Mike Scott

28 replies on “Government Gives £38 million Extra To A Mental Health Service That Has Already Crashed With Real Life Clients”

Hi Mike,

I really enjoy your work. I often look at this website to regain a sense of sanity from working at IAPT.

This piece in particular hits home for me. The fact that we are ill equipped (even when qualified) and that we are expected to deal with complex cases (when we are told that we aren’t supposed to….per se). It does feel similar to being at war with a system that will never change. Unless you have friends in high places, I am afraid things will remain the way they are.

I remember asking a supervisor outright “why are we required to deal with these complex cases? Can’t they go to psychiatrists or something instead of us who can only offer 4 flavours of homework?” And they replied “oh well, some fall in the cracks”, basically a get out clause statement without any intention of changing things. We are sometimes expected to deal with convicted (or worse, ongoing) sex offenders, abusers and so on. “It’s our job” they say at work but they never give us the respect of not wanting to work with particular patients out of personal grievences with them or just general bad personal experiences. Heck, we are expected to work with ex-veterens, which again, I feel would benefit from psychiatrists or someone better than “here is your homework for next week” schtick. I feel ex-veterans paid their due and they deserve something more than bare bones treatment. As someone once said, working at IAPT is like having all the responsibility without any of the authority to do anything. It’s never “how did it go with x patient?”, it’s always how much can we jam in your workload and then, what are your DNA, cancellation and recovery rates?

I keep coming back to your website because it reminds me that I need to not get comfortable and to not develop Stockholm Syndrome at IAPT. No disrespect to IAPT, but one has to be asleep in order to believe in it. I just wanted you to know that you give people like me the courage to focus on a career change, rather than “let sleeping dogs lie”. All i can say is, as soon as the reality set in of what it’s like working at IAPT, I felt like I had to be a salesman rather than a therapist. Selling them on ideas rather than actually helping them. Luckily, I have lived a life that has given me enough to talk about at helpful tangents with patients. Otherwise, the bare bones of what IAPT gives, is like knowing how to ride a bicycle, but being forced to do it blindfolded.

Also, as you have mentioned a lot in this website, we don’t give formal diagnosis. So this further makes the work itself meaningless. It further, dare I say, makes me feel like a loser as opposed to a “clinician” that we are dubbed as being. I enjoy hearing the stories that patients bring. However, I cannot shake off the nagging feeling that it is all for naught at the end.

Thank you for keeping this website updated.


I must say that I find this post disingenuous, for the following reasons:

1. You use the Hepgul et al. 2016 study to make a point that most patients in IAPT will have multiple disorders and thus will be too complex for “minimalist, little contact with a therapist other than to supervise the client’s use of computerised CBT”. You fail to point out that two thirds of the patients in Hepgul’s study were on the waiting list for a high intensity intervention (usually standard CBT). So your comment about self-help strategies found in low intensity IAPT work being of limited use doesn’t really apply here (not to mention that low intensity is much more than just computerised CBT).

2. How does Zimmerman et al. 2005, a study investigating psychiatric outpatients in Rhode Island (America), highlight the complexity of IAPT’s target patients?

3. Regarding your point about Hepgul et al. 2016 and history of abuse, are you insinuating that, by having a history of abuse (what does that even mean? How was it defined? How severe? What was the participants’ understanding of abuse? Was the trauma resolved with previous therapy? Did the participant decline any higher intensity support specific for the trauma?) a patient can therefore not benefit from any guided self-help for a specific presenting problem? What would the evidence for this be?

4. How do you know that IAPT training is not sufficient in teaching prevalence, complexity, or multiplicity of mental health disorders, or the ability to distinguish PTSD from a phobia? As a PWP I’ve had training on many of the things you say we don’t have training on. Is the training perfect? No. Does your average PWP know the difference between a situation bound panic attack and panic disorder? I’d be shocked if not.

5. Your study of 90 IAPT patients with a personal injury claim has absolutely no bearing on what recovery rates might look like in IAPT. That would be like doing a study of 90 Vodafone customers who have sued them for breach of contract and then using that to work out Vodafone’s customer satisfaction rate.

Whilst you occasionally raise interesting points on this blog, I think there is a duty to presenting research honestly, and evaluating them fairly.

Hi Sergio and Nathan

To take your points in turn Sergio;
1. The purpose of the Hepgul et al (2016) study was to determine the clinical characteristics of the IAPT population, using a standardised diagnostic interview. This had not been done before. The authors believed that they had fairly sampled this population. There is no published study that suggests that their sampling was at fault. On this basis they concluded ‘the results have revealed the complex nature of the patients seen within an urban IAPT service, with high rates of psychiatric comorbidity, bipolarity, childhood trauma and traits of personality disorder’. I was simply making the point that using low intensity interventions with this population was unlikely to bear fruit. There are no published studies of the efficacy of low intensity interventions with these populations. This is not to say that low intensity may not be effective for some problems but we are bereft of studies using blind independent raters, conducting standardised diagnostic interviews to demonstrate this.
2. It is difficult in a blog to be concise enough and in retrospect I should have explained my reference to the Zimmerman study more. The rates of various disorders identified in this study were much the same as in the Hepgul et al (2016) study, including those for borderline and antisocial personality disorder. This suggests that the IAPT population is much the same as psychiatric outpatients. Zimmerman found that overall almost a third of patients had a personality disorder this was therefore most likely to be also the case in the IAPT population.
3. Determining a history of abuse is not an easy matter but I think that Hepgul et al (2016) did as good a job as anyone, which is what I would expect with people like Nick Grey authoring the paper. The authors were simply saying that in a population were a significant minority have been abused treatment is going to be challenging. It was beyond the initial remit of Hepgul et al (2016) to monitor the trajectory in the long term, though this is an important matter, to do so would likely require extra funding.
4. There is no evidence that IAPT training improves outcome, the study by Liness et al published in Cognitive Therapy and Research found that measuring therapist competence on the Cognitive Therapy Rating Scale did not relate to outcome. If IAPT therapists were competent enough to make diagnosis the IAPT Manual would not state that IAPT workers do not make diagnoses.
5. Litigation is a common phenomenon amongst those who have expressed a trauma, thus in the Blanchard et al study of CBT for PTSD and sub-syndromal PTSd 60% were involved in litigation. Thus there is nothing unusual about a focus on litigants, As a medico-legal Expert one tracks the trajectory of patients symptoms, in my 2018 study I found that those who had been through IAPT before their trauma did no better as a group than those who went through IAPT afterwards. Further overall just 10% recovered i.e lost there diagnostic status.

Nathan, what you say is what I hear time and time again. I think it must be awful to be in a system does not engage in open debate. But it is no better than BABCP, were the current and past President conspired to reject my offering for CBT Today on IAPT bypassing the editor.

Best wishes


Thanks for your reply Mike.

1. You say that you are simply making the point that using low intensity interventions with this population was unlikely to bear fruit – you may be right, which is why two thirds of their sample was on the waiting list for high intensity therapy, so the study doesn’t demonstrate your point.

2. I think it’s extremely generous to say that the evidence suggests IAPT patients are going to be largely the same as Rhode Island’s psychiatric outpatients. Even if our healthcare systems were identical (which they are very much not), it would still be a huge extrapolation.

3. Yes treatment may indeed be more challenging if there has been a significant history of abuse, but again the study does not demonstrate your point. A history of abuse does not contraindicate guided self-help, and the evidence you cited doesn’t suggest any reason to think this, so what was the point exactly?

4. There is some limited evidence of therapist effects within PWPs and their achieved outcomes (Green et al 2014, Firth et al 2015), but that wasn’t the point anyway. You were saying that a PWP can’t recognise one thing from another, which is not true. As to why IAPT doesn’t do diagnoses, I suppose that’s a whole different story and included many different layers, but I don’t think it speaks solely to competence.

5. Agreed, but again that’s not my point; the point is that it is by no means a representative sample. Again you state that only 10% lost their diagnostic status, which I don’t question. I only state that it is in no way, shape or form generalisable.

Hi Sergio

You are right Sergio, there is a need to replicate my findings Scott (2018) because of a question mark over representativeness, But it does provide a window into IAPT’s performance, unfortunately IAPT have taken no steps towards independent audit of their work. It has got away with marking its’ own homework by superb marketing and friends in high places, NHS England’s National Mental Health Director, Claire Murdoch in the last few days declaring ‘our world-leading talking therapies’, she would hopefully not say this of a drug that had not been independently evaluated by those other than the manufacturers of the drug .
The prevalence estimates for alll the common psychological disorders in the Zimmerman study were the the same as in the Hepgul study but additionally the former assessed for all the DSM personality disorders thus the likely incidence of PDs in IAPT is the same, There may be differences in income, etc between the 2 populations making the IAPT clients even more difficult to treat.
There is no single study to demonstrate that low intensity works with those that have been abused, using the Tolin criteria for an empirically supported treatment. There is no evidence that low intensity interventions leads to a loss of diagnostic status.
best wishes

Hi Mike, thanks for your reply. I agree with your point that your findings would need to be replicated and also that there should be an independent evaluation of IAPT.

I think the point about low intensity CBT for patients with history of abuse is complex; for starters, my point still stands that most of Hepgul’s sample were on the waiting list for HIGH intensity CBT. Secondly, would there be a strong rationale for hypothesising that patients with a brief history of small-scale abuse would not benefit from low intensity work (since we are very unlikely to work with clients with long history of substantial abuse)? I think it would be very interesting to study, but I think it’s inaccurate to assume that low intensity would have no benefits; the rationale for this is weak and Hepgul doesn’t support that rationale anyway.

I also agree with you that it would be interesting to study diagnostic status of IAPT patients. I do believe that recovery rate evidence (which I see firsthand in my service and caseload) is encouraging, but I do fully agree with you that there should be independent analyses as well. Notwithstanding concerns re: the usefulness of psychiatric diagnoses in the first place.

… [Trackback]

[…] Read More on to that Topic: […]

… [Trackback]

[…] Here you can find 52371 more Information to that Topic: […]

… [Trackback]

[…] Find More here on that Topic: […]

… [Trackback]

[…] Find More on that Topic: […]

Leave a Reply

Your email address will not be published. Required fields are marked *