[Today’s blog on the potential harms of psychotherapy is a joint effort. The first half (introduction, methods, results and reflections) was written by Keith Laws and the second half (next steps and J’Accuse) by Samei Huda.]

The emphasis on establishing the efficacy of psychological therapies has historically consigned harms to a largely unchartered hinterland. Information about both the risks as well as benefits, however, is vital if patients are to make informed choices about whether to engage in psychological treatment; in line with the NHS ambition of ‘No decision about me without me’.

We have quite recently seen a publicly funded trial of a psychological intervention halted because of worryingly higher rates of adverse events (overdose, self-harm and suicide attempts) in the intervention arm of the trial (McMurran et al 2011).  Two quite different recent reviews (Vaughan et al, 2014; Jonsson et al, 2014) have confirmed the low rates at which adverse events are monitored in randomised controlled trials of psychotherapy.

Most recently, a review of the final reports from National Institute of Health Research (the main public funder of psychological therapy trials in the UK), by Duggan and colleagues (2014) found that not one trial of psychological therapy trial funded between 1995 and 2013 mentioned the occurrence of an adverse event (compared to over two-thirds of RCTs involving drug treatments).

Methods

The current paper by Crawford and colleagues (2016) adds important novel information to this sketchy and perturbing picture. Crucially they conducted a real-world survey of the negative experiences reported by patients. The National Audit of Psychological Therapies is a large-scale assessment of state-funded psychological therapy services for adults with depression and anxiety in England and Wales. The audit examined clinical records and surveyed people in both primary and secondary care services and approximately 60% (220 services) of all eligible NHS services participated.

Patients aged over 18 and receiving out-patient treatment for anxiety and/or depression were invited to complete an anonymous (online or paper) questionnaire. All who participated expressed whether they had experienced ‘lasting bad effects from the treatment’. The terms ‘lasting’ and ‘bad effects’ were left quite opaque.

Results

Of the large sample of 14,587 participants surveyed:

• 763 (5.2%) slightly or strongly agreed they had experienced lasting bad effects;
• and an additional 1,099 (7.7%) were unsure.

Crawford and colleagues also assessed how various demographic and clinical factors were associated with the experience of lasting bad effects:

• Most patients reported being referred at the right time, offered the right number of sessions and given sufficient information about treatment before it commenced
• Lower reporting of negative effects was associated with various pragmatic variables including whether patients were able to get to appointments without difficulty and whether they were scheduled at convenient times
• Patients aged over 65 (and especially those aged over 75) were less likely to report negative effects than younger patients
• The most striking findings, however, were the much higher rates of long-lasting negative effects of psychological therapy reported both by people from Black and Minority Ethnic groups (BME) and by non-heterosexuals (Lesbian, Gay and Bisexual: LGB).

Reflections

Although a figure of 5.2% reporting lasting harm may seem quite small, it needs to be viewed in context. In England alone, the past few years have seen over one million people referred for psychological therapy for common mental disorders, suggesting that as many as 50,000 may have experienced lasting bad effects; and many more left ‘unsure’.

The findings regarding BME and LGB groups are especially worrying. We know that LGB adults have a higher prevalence of poor mental health and lower wellbeing when compared to heterosexuals (Semlyen et al 2016) and this study suggests they are also at greater risk of long-term negative effects following psychotherapy. Similarly, it has been long-known that people from minority ethnic groups experience “persistent inequalities in both experiences and outcomes” in their contact with psychiatric services (Bhui et al 2015) and that UK clinical psychology services “…are failing to meet the psychological and clinical needs of people from minority ethnic groups” (Williams et al 2006).  The current survey leads us to the worrying notion that individuals already at increased risk of mental health problems may be more likely to report negative experiences resulting from psychotherapy.

As noted in an accompanying editorial by Scott & Young (2016), the paper by Crawford et al overlooks one extremely large elephant-in-the-therapeutic room -how do therapist characteristics contribute to the negative experiences of therapy? The importance of therapeutic alliance (TA) to outcome has been a consistent theme since the time of Freud and requires more research attention.

It is also worth noting that Crawford et al examined only outpatients diagnosed with anxiety and/or depression. Future studies would need to look at rates of negative experience amongst those with more severe mental disorders. It would also be prudent to investigate the issue of negative effects in younger people, as Parry et al note in the accompanying editorial “negative therapeutic results may be higher for children and young people undergoing psychological therapy”. The study also assesses patients within the NHS and future work might explore the comparable figures in private practice.

We should also be mindful that, like medication, psychological therapies are something of a curate’s egg. As Parry et al (2016) rightly point out in their accompanying editorial piece “it is becoming clear that psychological treatments cannot be at once psychoactive and harmless”. One individual may experience both beneficial and harmful consequences in different areas of their life. Harms and benefits are no more mutually exclusive than they are with medication. The current lack of attention to harms resulting from psychological interventions means that patients are not able to make informed decisions about treatment. The recognition of potential harms has been a remarkably slow, but studies such as this provide crucial information about the experiences of patients.

[Thanks Keith! Over to you Samei…]

Next steps

As a clinician, should I be warning my patients that if they belong to BME or LGB(T) groups, they have twice the risk of lasting bad effects from psychotherapy? Or should I take into account the flaws in this study, but also remember the reasonably large numbers they did manage to get information from?

At the very least, psychotherapy researchers need to rapidly investigate if this study was a “chance” result or has discovered a significant difference in risk of harm after psychotherapy in some ethnic and sexual orientation groups. If this increased risk from psychotherapy for certain groups is replicated and confirmed, then researchers should investigate if four types of broad explanations are to blame (they may all be present to varying degrees):

1. Participants referred for psychotherapy
2. Psychotherapy as a treatment model
3. Psychotherapists
4. The broader contexts in which psychotherapy is delivered

Taking the first possible explanation, about 84% of participants were referred for psychotherapy by professionals such as GPs. They may be inadvertently referring people from ethnic minorities and sexual orientation minorities who are at particular risk of harm from psychotherapy.

Psychotherapy as a treatment model may be more likely to cause harmful effects in certain ethnic groups and minority sexual orientations. We know that certain medicines may be less effective or more likely to cause harmful effects in certain ethnic groups. Perhaps this may apply to psychotherapy (or certain types of psychotherapy) too, but for different reasons.

In psychotherapy, the relationship between the therapist and the client is crucial. It is possible that some therapists are less able to form positive therapeutic relationships with the ethnic and sexual orientation minority groups described above, leading to increased risk of harmful effects. This could potentially be overcome if clients from these minority groups saw therapists from the same minority group as them.

The type of problems that Asian/Chinese/Black/bisexual/other sexual orientation people present with seeking help from psychotherapy may not be best served by the way psychotherapy is delivered to them by the NHS. The pressures IAPT faces are well documented. The delivery of psychotherapy by the NHS may increase the risk of harm to the minority groups described above.

These suggestions I have made are clearly speculative and would need to be examined in a good quality research study. And that’s where the problems begin.

J’accuse

The problem with a great deal of psychotherapy research is that it neglects the issue of harm caused by psychotherapy. There is a lack of a standardised terminology to classify and measure these harms (though an accompanying editorial does make some suggestions).

There are some researchers working on this important topic (such as Glenys Parry), but we need qualitative research to identify the different types of harms experienced as a result of psychotherapy and to try and understand the processes leading to it. We then need the development of research methods such as ways of measuring the frequency and severity of these different types of harm identified by qualitative research. We can then move onto properly assessing the extent of harm caused by psychotherapy and how long it persists.

That this has not happened already is puzzling. Psychotherapists try and minimise harm to their clients by therapy through improving their skills by training and discussing cases in supervision. Yet this has not seemed a priority for researchers. Their focus seems to be on proving the effectiveness of therapy and the processes underpinning this.

This reflects a focus on the interests of psychotherapy and psychotherapists but not that of their clients. Clients clearly benefit from evidence of effectiveness but they also benefit from research on harms, which allows them to make informed choices and it helps the therapist minimise further the risk of harms. I can think of few types of treatments outside alternative medicine where there is such poor systematic collection of harmful effects caused by the intervention during research studies, barring details of self-harm and sometimes symptom ratings.

It is time to put the interests of clients first, by seriously researching the harms of psychotherapy. It should also help us understand if the result of this study is true, and what to do about it.

Links

Primary paper

Crawford MJ, Thana L, Farquharson L, Palmer L, Hancock E, Bassett P, Clarke J, Parry GD. (2016) Patient experience of negative effects of psychological treatment: results of a national survey. The British Journal of Psychiatry Mar 2016, 208 (3) 260-265; DOI: 10.1192/bjp.bp.114.162628 [Abstract]

Other references

Berk M, Parker G. (2009) The elephant on the couch: side-effects of psychotherapy. Australian and New Zealand Journal of Psychiatry, 43(9), 787-794.

Benish SG, Quintana S, Wampold BE. (2011) Culturally adapted psychotherapy and the legitimacy of myth: a direct-comparison meta-analysis. Journal of Counseling Psychology, 58(3), 279. [PubMed abstract]

Bhui K, Aslam RW, Palinski A, McCabe R, Johnson MR, Weich S, et al. (2015) Interventions designed to improve therapeutic communications between black and minority ethnic people and professionals working in psychiatric services: a systematic review of the evidence for their effectiveness. Health Technol Assess 2015; 19(31).

Dimidjian S, Hollon SD. How would we know if psychotherapy were harmful? Am Psychol 2010; 65: 21–33. [PubMed abstract]

Duggan C, Parry G, McMurran M, Davidson K, Dennis J. (2014) The recording of adverse events from psychological treatments in clinical trials: evidence from a review of NIHR-funded trials. Trials, 15(1), 335.

Jonsson U, Alaie I, Parling T, Arnberg FK. (2014) Reporting of harms in randomized controlled trials of psychological interventions for mental and behavioral disorders: A review of current practice. Contemp Clin Trials. 2014 Mar 4;38(1):1-8. doi: 10.1016/j.cct.2014.02.005. [Epub ahead of print] [PubMed abstract] [Accepted manuscript PDF]

McMurran M, Crawford MJ, Reilly JG, McCrone P, Moran P, Williams H, Adams CE, Duggan C, Delport J, Whitham D, Day F Psycho-education with problem solving (PEPS) therapy for adults with personality disorder: a pragmatic multi-site community-based randomised clinical trial. Trials. 2011 Aug 24; 12():198.

Reuter L, Bengel J, Scheidt CE. Non-response to therapy in acute and rehabilitative psychosomatic inpatient care – a systematic review. Z Psychosom Med Psychother 2014; 60: 121–45 [Abstract]

Scott J, Young AH. (2016) Psychotherapies should be assessed for both benefit and harm. The British Journal of Psychiatry, 208(3), 208-209. [Abstract]

Semlyen J, King M, Varney J, Hagger-Johnson G. (2016) Sexual orientation and symptoms of common mental disorder or low wellbeing: combined meta-analysis of 12 UK population health surveys. BMC Psychiatry, 16(1), 1-9.

Vaughan B et al. (2014) Frequency of reporting of adverse events in randomized controlled trials of psychotherapy vs. psychopharmacotherapy. Compr Psychiatry. 2014 Jan 23. pii: S0010-440X(14)00006-6. doi: 10.1016/j.comppsych.2014.01.001. [Epub ahead of print] [PubMed abstract]

Williams PE, Turpin G, Hardy G. (2006) Clinical psychology service provision and ethnic diversity within the UK: a review of the literature. Clinical Psychology & Psychotherapy 13.5 (2006): 324-338. [Abstract]

Psychotherapy trials should report the side effects of treatment

Photo credits

• Erich Ferdinand CC BY 2.0
• Rick&Brenda Beerhorst CC BY 2.0
• R/DV/RS CC BY 2.0