In NHS Talking Therapies We Doubt


This is the take home message of a just published study in the British Journal of Clinical Psychology. Capobianco et al (2023) The authors conclude “Significant questions must remain concerning how effective the treatment element actually is and how much time is needed to realise an adequate dose”. With mean improvements of 2-3 points on the PHQ-9 between 1st and last session and a mean 3 point improvement on the GAD-7 between 1st and last appointment, whether or not treatment was conducted remotely or face-to-face. They further add “however, we are not arguing the treatment was effective and therefore the cause of the changes observed. Such changes could be accounted for by a range/ combination of factors including regression to the mean or spontaneous recovery over time…….. It seems that clinical improvement was slow, and patients do not appear to be receiving the required length of time in order for outcomes to reach the required cut of”.

Their data reflect the failure of NHS Talking Therapies to engage clients thus, “between March 2020 and September 2020, 5515 patients attended at least one session, with 2553 (46%) patients attending at least two treatment sessions.  Similarly pre Covid19, 9199 patients attended at least one session, with 4625 patients (46%) receiving at least two treatment sessions. Participants attended a median of two remote therapy sessions and a median of three in person therapy sessions’. By anybody’s reckoning the median dosage of therapy is sub-therapeutic.

Dr Mike Scott

152 replies on “In NHS Talking Therapies We Doubt”

I am a CBT therapist with IAPT. It certainly has flaws, and there is certainly a case for independent review. However I feel your argument somewhat misrepresents the service in one major aspect: many clients who attend only one appointment are referred on eg to CMHT, to psychotherapy, to specialist agencies such as domestic violence charities, LGBTF 3rd sector, specialist trauma provision, rape crisis etc. In our area clients tend to access 1st assessment via external non nhs step 2 agency (PWPs), a proportion are then referred to our nhs step 3 service (CBT or counselling, with patchy additional access to EMDR, CFT, DIT, etc). It’s a gateway sorting pyramid. In my view the cycle of multiple waits/assessments is frustrating and damaging and needs review. But your figures need more context to serve the argument better – admittedly hard when IAPT lacks transparency and when services are driven to meet targets which encourages dubious practices.

Hi Paula
Unfortunately the idea that IAPT appropriately signposts one session attendees (half of those who enter the Service) and the failings are restricted to those who engage in treatment ( i.e engage in two or more sessions), lacks credibility. In 50-60% of cases in the study no diagnosis was made, there is no indication that this was any different for one session attendees, so that for over half the sample there can be no confidence that they were placed on the right referral pathway. The mean initial PHQ-9 score of the whole sample was 16 with a standard deviation of 6, this means that 68% of the sample would have been judged a case of depression, ( a score above 10) but only 26% of the sample were given a label of depression! Informed by these diverging metrics the clients could be sent anywhere. It could be argued that the PWPs, who most likely conducted the assessment, have some special knowledge about putting people on the right pathway. But such ‘special knowledge’ has not been demonstrated.
Best wishes


I agree Mike that there is a problem in the fact that those with the least training are effectively the gate-keepers. In practice this means at each assessment at whatever step the assessor asks “does this look too complex for me and the interventions I have available” and if so passes it up the food chain. This is hugely inefficient as complex patients are routinely being assessed 3or4 times (for an hour each time) before any kind of treatment, with a wait of at least several months between each onward referral. I am not sure what the solution is though – could a large enough pool of highly trained assessors be created realistically?
Either way, I am glad there are critical eyes like yours on IAPT – I think it can give great benefits to many patients, but it does seem to have become anathema to query its workings and figures.

Hi Paula
I think your idea of a pool of highly trained assessors is fantastic! But when I suggested taking much more time over assessments using a ‘gold standard’ diagnostic interview David Clark’s response was that he couldn’t see himself being able to get it through the Commissioners. The difficulty is moving the debate beyond numbers and waiting times and really listening to clients
Best wishes


Leave a Reply

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