IAPT Has Been A Law Unto Itself and is Rightfully Being Ignored by NICE

The National Institute for Health and Clinical Excellence (NICE)  can only base its treatment recommendations on studies that have a rigorous methodology. In generating the proposed recommendations on the treatment of depression &source=web&cd=&ved=2ahUKEwic8KOfmcD1AhVOasAKHVt8C_EQFnoECAcQAQ& NICE has ignored all studies that  emanate from the Improving Access to Psychological Therapies (IAPT). Yet the lead organisation for cognitive behaviour therapy, the British Association for Behavioural and Cognitive Psychotherapy (BABCP) BABCP response – NICE consultation draft has protested vehemently about this. But applying the ‘Psychotherapy outcome study methodology rating form’ developed by Ost (2008) OST the original randomised controlled trials of CBT for depression and anxiety disorders had a mean score  of 27.8, (SD 4.2). Applying the scale to studies by IAPT related personnel, they struggle to score into double figures – a fate shared by studies of low intensity CBT.  To put these scores in context, Ost (2008) found the total mean score for ACT was 18.1 (SD 5.0) and for DBT 19.4 (SD 3.9). He considered the scores for ACT and BDBT too low, for them to be regarded as Evidence Supported Treatments (ESTs). How much less of an EST then are the IAPT interventions? BABCP defends itself by saying the IAPT studies need to be evaluated by some other metric, but don’t specify which. This sounds suspiciously like the defending of a family member, rather than being data driven.

It should be noted that IAPT does not measure either adherence  (item 15 on the rating form) nor competence (item 16 on the rating scale). Thus there is no assurance of treatment integrity in IAPT. IAPT clinicians have been a law unto themselves. NICE therefore cannot be sure that IAPT’s alleged treatment interventions were delivered.

The studies by IAPT related personnel fail abysmally on almost every index of reliable methodology. Running through the rating form: IAPT therapists do not make diagnoses, making for ‘0’ scores for items 1-6, similarly ‘0s’ would be awarded for no blind evaluators (item 7), no assessor training (item8), no random, assignment to treatments (item 9), no control groups (item 10), treatment as usual (item 11), no power analysis (item 12), only pre and post assessment points  (item 12), effects of therapist were not assessed, nor level of training  [items 14 & 15generously a score of 1 could be awarded on both these items, no control of concomitants (item 18), no intention to treat analyses (item 19), statistical analysis confined to completes (item 20),  no evidence of real world clinical significance (item 21 but a case could be made for awarding a 1 score, no equality of therapy hours because no comparison condition (item 22).  

The low intensity rcts similarly rate very poorly on the rating form. Studies of these cheap offerings rely on establishing statistically significant differences with a comparison group. Never stopping to assess whether any found difference is clinically  meaningful. Any differences do not pass the ‘Does it matter? test, or the ‘So what? test or the ‘Why should anyone care?’ test. In none of the studies have clients been asked independently post treatment  ‘are you back to your usual self, now?’ Importantly if they reply ‘yes’,  then asking ‘for how long have you been back to your usual self?. Studies of the natural history of anxiety disorders have utilised a period of 8 weeks free of  meeting diagnostic criteria, to define recovery, Bruce et al (2005)] The absence of data on the proportion of clients returned to their normal and enduring functioning  by these ‘cost-saving’ interventions, means that prospective clients cannot make an informed choice about engaging in such treatments. NICE needs to proceed more cautiously in recommending low intensity CBT.


Dr Mike Scott