BACPs latest public attitudes survey reveals that 77% of people who’ve had therapy would recommend it to somebody who had emotional difficulties or a mental health problem. Question is, should we be throwing our hats in the air at that percentage, or should we be concerned?
Estimated reading time: 4 minutes
Table of contents
77% of respondents would recommend counselling
Following a recent tweet from BACP, I’m trying to work out whether responses to BACPs latest public attitudes survey represent good news or bad news, or just a mixed and rather confusing picture.
The survey of 5,333 UK adults was carried out online between 9 to 22 February 2023. Among its findings were the following:
- 95% of people who’ve had counselling or psychotherapy think it’s important it should be accessible to everyone who wants it.
- 92% of people who’ve had counselling or psychotherapy agree it’s a good idea to seek counselling or psychotherapy for a problem before it gets out of hand.
- 82% of people who’ve had counselling or psychotherapy agree people might be happier if they talk to a counsellor or psychotherapist about their problems.
- 77% of people who’ve had counselling or psychotherapy would be likely to recommend it to somebody who had emotional difficulties or a mental health problem.
So, the principle that therapy should be accessible to anyone who wants it gets the highest level of endorsement. It’s followed by sentiments that it’s a good idea to tackle problems before they escalate, and that people might be happier if they did.
The lowest level of endorsement, however, is from the 77% who would be likely to recommend therapy to someone who might need it. If this were a 4-way medal competition, actual recommendation of therapy from those who’ve received it wouldn’t even get a place on the podium.
Hardly a ringing endorsement…..
I don’t know about you, but if I’m buying goods or services I’m not familiar with, I’m likely to turn to some form of consumer rating. Generally, anything with less than four stars tends to get a wide berth.
On the basis that 77% of previous consumers of therapy would recommend it to others, I think there’s grounds for arguing that we have something of an image problem.
Here’s the survey finding again:
“77% of people who’ve had counselling or psychotherapy would be likely to recommend it to somebody who had emotional difficulties or a mental health problem”
Is that percentage something we should be content with, or not? If you think that statement is a positive endorsement, let me offer you a reframe.
“Nearly one quarter of respondents who’ve received therapy didn’t say they would recommend it.”
However the statement is phrased, to my ears it hardly sounds like a ringing endorsement. And, given that it comes from people who have previous experience of therapy, what is it saying about the experiences of the missing 23%?
Are we failing the 23 percent?
We know, overall, that therapy works. With an effect size of around 0.8, therapy has a large effect. The average ‘treated’ client is better off, psychologically speaking, than 79% of people who are untreated.
But we also know that therapy’s effects aren’t uniform. Not everyone benefits. Impact varies across therapists, services and settings. We’re not all the same.
I’ll be happy when the proportion of people who’ve had therapy and would recommend it reaches 100%. If we’re to shift the dial on public perception, however, I think we need to do a little better.
Attrition rates in some settings are inexplicably high and improvement rates worryingly low. IAPT is a prime example. As the graphic below shows, a significant amount of clients don’t go beyond a first session. Of those that do, a significant proportion fail to demonstrate improvement.
I’d dearly love to know what therapy the 23% who didn’t recommend therapy received, and what their experience of it was. I’d dearly love to be able to drill down and see what we might learn about their experience.
Sadly, that opportunity has gone But it really needn’t stop us from working hard to understand the experience of clients who we’re working with right now, as well as those in the future. Let’s do what we can to all try and get a little closer to that 100%.
Just like you we thrive on feedback.
Please leave your thoughts on what you’ve read in the comments section below.
Another good request for less games playing and more attention to detail: thanks Barry.
But just to be my usual obsessional self, you say:
“We know, overall, that therapy works. With an effect size of around 0.8, therapy has a large effect. The average ‘treated’ client is better off, psychologically speaking, than 79% of people who are untreated.”
That could read that post-therapy, the average client is in the top 20% of the population. I don’t think that’s what that effect size means. No citation to your source for that so I can’t be sure, but I think it means that, compared to an untreated comparison group with similar problems and seeking therapy, the ones who received therapy end up with scores on whatever scale is being used, at the point at which that group finish therapy, that are better than 79% of those who didn’t get therapy. That’s better than the roughly “better than 50%” that you’d expect if therapy had no good effect and it’s not bad (though of course it hides a lot of diversity in changes) but it’s not what I think your two sentences imply.
Of course, my other caveat about these sorts of effect size reports is that they’re often on very short term follow-up or simply about scores at the last session of therapy. Again, that’s better than nothing but when many people, particularly those with physical health and/or socio-economic problems may have long term or recurring psychological needs, it’s a very partial answer. We still need a fundamental change in how we research and assess therapies. Imagine if cancer, or diabetic crises were only assessed on immediate post-treatment state. Both are about long term prognoses and both mortality and morbidity. If MH/WB/therapies were taken seriously and funded seriously we wouldn’t be given largely silly statistics, whether by BACP (sorry BACP) or by IAPT and the NHS and you could have a happier job with TMN! Very best, Chris
Hi Chris, and thanks posting!
Your usual obsessional self is very welcome and keeps me on my toes! I did put in a link to the source data, which is The Great Psychotherapy Debate: The Evidence for What Makes Psychotherapy Work by Bruce Wampold and Zac Imel. With data that can be so contested, maybe I would have been better putting in a citation! This is taken verbatim from p94:
“A reasonable and defensible point estimate for the efficacy of psychotherapy would be .80 , a value used in this book. As indicated in Table 3.1, this effect would be classified as a large effect in the social sciences which means that the average client receiving therapy would be better off than 79% of untreated clients, that psychotherapy accounts for about 14% of the variance in outcomes…………..”
So, yes, I think that equates to your interpretation above “but I think it means that, compared to an untreated comparison group with similar problems and seeking therapy, the ones who received therapy end up with scores on whatever scale is being used, at the point at which that group finish therapy, that are better than 79% of those who didn’t get therapy.”
On your other points – I could not agree more! As for MH/WB/therapies being taken seriously and funded seriously – I look forward to that happy day (though I may not be around to see it!)
Cheers! Barry
Hi Barry. I agree with all this – and of course there is another important angle that wasn’t asked about at all – how many people would say that therapy actually made things worse? This would seem to be a distinct possibility in IAPT, where recovery rates seem to be lower than the expected natural rates of recovery over the same period.
Please can you clarify what is meant by ‘ended therapy’ in the chart? This is a very badly-worded heading. Does it mean that they ‘completed’ their therapy? (which they didn’t necessarily, if the criterion includes attending a minimum of 2 sessions). Or does it mean the opposite – ie that the client ‘ended their therapy’ – ie decided not to continue with it, for whatever reason?
Hi Lucy and thanks for posting. How many people would say that therapy actually made things worse – wouldn’t we like to know? And it begs the question of how it’s measured or assessed and by whom? I’m not sure if you looked at the blog from which the chart is taken, but basically all those coded as ended treatment have attended two or more appointments. This includes both endings that were planned as well as those that finished prematurely. So it’s a catch all that doesn’t distinguish further and I would agree it’s rather opaque and open to interpretation. Over the years I’ve found IAPT data to be less and less available and comprehensible – less transparent basically.
Flipping it on its head – maybe it’s less about the Counselling but more to do with the question, and in particular the ‘somebody’ that respondents were thinking about when answering this question? Maybe they don’t think that the ‘somebody’ they had in mind would engage; have the ability to self-reflect; or be motivated to make changes? Their thinking might be defaulting to a stubborn partner or a rude boss – who may even be at the core of the respondent’s own issues? Just a thought…
Still – it would have been great if a follow-up question were added to this survey e.g. ‘If you are not likely to recommend therapy, please comment.’
Hi Elsa – thank you, and I so agree with your final point about the follow up question. Now that really would give us something to work with. Maybe it just flies too much in the face of the broad intentions of a public attitudes survey though…………sadly?
Thank you for this incisive article. Two comments:
1) might want to downsize the ES of .8, given the trend towards diminishing ES observed over the last 15 years of meta-analytic reviews (e.g., anxiety/ depressive outcomes.) It seems the closer we look, the more sobering the findings.
2) given the evidence for inadvertent iatrogenic harms arising from psychotherapy, I’m thinking continuous therapy process/outcomes measurement / management has become a sine qua non for clinical practice.
Qui custodiet ipsos custodes?
Yup. Us.
Respectfully
Hi David and thanks for your post – ES is such a moving target isn’t it. I think Chris Evans’ comments on this post are also very relevant, and I’ve expanded on the source material in my reply. I want to say “if it’s good enough for Bruce Wampold it’s good enough for me” but other perspectives are available and worthy of our consideration….. Thanks again!