Clinical Psychology Forum 162 - June 2006


The trouble with psychotherapy


Paul Moloney


The supposedly sound evidence base for the effectiveness
of the psychological therapies may be far more
questionable than is widely supposed.

The term `psychotherapy' refers to a
broad family of talking treatments for
personal distress which are of proven
effectiveness, where some approaches are
helpful for certain kinds of personal difficulty,
and where accredited professional
training will confer particular expertise and
skill (see McLeod, 1994).

The assumptions
contained within this statement are unlikely
to meet with dissent from the average person
in the street (see Furedi, 2003), and in one
form or another constitute the taken for
granted world of the psychotherapy professionals
themselves.

For example, they are
endorsed in the training programmes of
clinical and counselling psychology (BPS,
2002, 2004), in central government recommendations
for the use of psychological therapies
in the NHS (DoH, 2001) and most
recently in calls by Richard Layard - one of
the UK government's key economic consultants
- for psychological therapy to be made
`available to all', as the main answer to the
personal and social malaise which seems to
be afflicting us at record levels (see Roth &
Stirling, 2006).

In contrast to this warmly consensual
picture, the enduring reality is that the psychotherapy
outcome literature offers precious
little support for any of the above
notions.

This is an observation that surely
has some importance for any profession that
concerns itself with the understanding and
alleviation of personal distress, and yet it is
one that seems to have been consistently
ignored.

In this paper I critically review some of
the key psychotherapy outcome literature and
ask why it has been so hard for psychologists
to acknowledge the poor evidence on therapy
outcome.

Does therapy work?

It is widely recognised that there are numerous
and complex difficulties in assessing the
effects of psychological therapies, and that
one of the best ways of doing this is by means
of the randomised placebo control trial (or
RPCT).

A large number of such investigations
have been conducted over the last halfcentury
or more, and although the results of
these studies have often been extremely variable,
so called meta-analyses - in which the
findings of large numbers of studies are
aggregated and then analysed - suggest that
most forms of psychological therapy are at
least mildly helpful.

Effect sizes upward of
half a standard deviation or more are routinely
touted.

That is to say, undergoing psychological
therapy is claimed to reliably lead
to significant improvement in the mental
health of up to a quarter or above of all
recipients (Smith et al., 1980).

This compares
favourably with other psychiatric treatments
that may themselves have a large
placebo component, such as antidepressant
medication (Breggin & Cohen, 1999).

Not surprisingly perhaps, these claims
seem to be authorised by the core psychotherapy
professions (see Fonagy & Roth,
1996).

Yet there are a number of serious
methodological problems associated with
attempts to assess the effectiveness of counselling
or psychotherapy in this way.

So much
so that at least some academics and practitioners
admit that it is hard to decide
whether these studies as a whole do or do
not support the notion that psychotherapy
or counselling is generally helpful (McLeod,
1994).

To begin with, the field has long suffered
from a bias toward the selective reporting
and publication of those studies that show
only the desired positive results (Boyle, 2002).

Many psychotherapy RPCT trials have
included inadequate control groups for
comparison purposes, often consisting of
individuals who remain on a waiting list or
who receive a less credible form of pseudotherapy,
delivered with visibly limited commitment
by the researchers (Holmes, 2002;
Mair, 1992).

Conversely, there has been a
trend toward excessive reliance upon
selected research populations, such as university
students or individuals with less severe
problems than are typically found in clinical
settings.

A large proportion of studies have
also suffered from systematic participant
attrition or selection effects that make the
results hard to interpret (Dineen, 1999;
Eisner, 2000).

Statistically significant differences
in outcome between participant
groups have often concealed large numbers
of people for whom psychotherapy has been
ineffective, while assessments of outcome
have tended to use abstract numerical measurements
and preset diagnostic inventories
that leave little room for subjective experience,
and which may therefore have limited
personal or even clinical meaning (Kline,
1988, and see Tolman, 1994),

Aside from these far from minor difficulties,
this literature may suffer from an even
more pervasive problem.

This is the tendency
to rely almost exclusively upon the
reports of participants - including the client,
the clinician and workers from the agencies
and institutions that support the therapeutic
work - in the absence of any fully independent
check upon the treated person's progress
in the world outside the consulting room
( Eisner, 2000; Epstein, 1996).

This is a serious
issue in psychotherapy research, because of
the range of powerful social and interpersonal
influences are likely to be in play, in
what is in many ways a unique situation in
our culture: part confessional, part ritual of
healing and social affirmation, and m u c h
else besides (Frank & Frank, 1991).

On this
basis, it may be worth discussing the question
of bias in client reports in more detail.

To start at the most basic level, both c l i e n t
and clinician will from the outset usually
desire the same broad result: an improvement
improvement
in the former's mental health, whether
this is defined as happiness, adjustment or
relative freedom from distress.

The patient's
cooperation towards this aim will be engaged
through the practitioner's efforts to establish
a therapeutic relationship, which implicitly
entrains the client into the given therapeutic
model.

For instance, many humanist therapists
seek to build a relationship with their
client that is intense and deep enough to
exceed most ordinary professional-layperson
encounters (e.g. Mearns, 1994).

Other practitioners
may emphasise the complexities of
the psychodynamic transference relationship
(e.g. Casement, 1995) or the alleged scientific
and technical authority for what they do - as
in cognitive behavioural therapy (e.g. Hawton
et al., 1989).

As a result, the client will likely have invested
a great deal of trust and hope in the
person of the practitioner.

All the more so,
perhaps, for having disclosed worries and fears
hitherto shared with few others.

Both parties
may also share potent, culturally sanctioned
beliefs, which equate failure to benefit from
therapy with the client's wilful rescinding of
the inner strength or discipline needed to
overcome adversity (Cushman, 1995).

For
the latter, these factors may combine to render
any admission of failure in the therapeutic
process a sign of personal inadequacy and
a source of anxiety about earning the tacit
disapproval of their therapist.

It therefore
seems reasonable to think that such admissions
of disappointment might be avoided or
denied by many clients: even - or perhaps
especially - to themselves (Epstein, 2006;
Kline, 1988).

And indeed at least some clinicians
felt that they have observed this process
in action (Kelly, 2000).

The key question, of course, is to what
extent the claims for psychotherapy effectiveness
might be distorted by this shaping of
client self-report.

As William Epstein points
out, the scale of this problem becomes apparent
when estimates of psychotherapy effect
size are compared with estimates of so-called
`demand characteristics'.

These are the expectations
that researchers can unconsciously
convey to participants in laboratory based
psychological experiments.

In the absence of
thorough controls such demand factors can
typically account for between 0.70 and 1.0
standard deviation of the reported effect
sizes.

This is for situations that are relatively
impersonal and short term in comparison to
most psychotherapeutic interventions, and
in which the participants might therefore be
expected to have a much lower stake in the
final outcome (see Rosenthal & Rubin, 1978).

Nevertheless, Epstein notes that these estimates
of researcher influences at least equal
(and often surpass) the average gains reported
for psychotherapy, even for the bettercontrolled
studies.

This is obviously a basic issue for the psychotherapy
field, where therapist expectations
of client improvement are inbuilt for
virtually every approach.

Yet rather than getting
to grips with these findings, the whole
area seems instead to have continued to rely
upon the reports of clients (and other
closely involved parties) in the absence of
any form of investigation that is external to
the therapeutic process or the organisations
that sponsor it.

The result being that
researcher expectancy cues are inseparable
from virtually all of the RPCT research to
date, and may confound it.

In the end, the
clear possibility remains that most of the
claimed benefits of psychotherapy might reside
in placebo effects (Epstein, 2006, 1996).

This last prospect is strongly underscored
by four further lines of evidence.

These are,
first, that, aside from (decidedly) modest indications
for the greater efficacy of behavioural
approaches in relation to phobias, the
comparative research literature seems to
offer little support for the idea that any one
treatment is more effective than another
(Assay & Lambert, 1999).

This observation
seems hard to reconcile with the confident
assertions of therapeutic potency and specificity
that are often trumpeted by adherents
of the mainstream therapies (Hansen et al.,
2003).

Yet, within the field, `there is tremendous
resistance to accepting this finding
as a legitimate one' (Bergin & Garfield, 1994,
p.822).

Second, the available evidence suggests
that, rather than specific techniques, a range
of so-called `non-specific' factors may account
for most of the beneficial effects of psychological
treatments.

Among these features,
the client's wider life circumstances and the
quality of the therapeutic relationship seem
to be the most important by far (Bohart, 2000;
Bergin & Garfield, 1994; Mahrer, 1998).

Third, comparisons of qualified practitioners
with amateurs who have received no
specific training in therapeutic models or
methods suggest that there are few real differences
between them in effectiveness, however
this is measured.

This is a surprisingly
robust (though, again, seldom acknowledged)
finding, which is supported by 39 separate
research studies conducted over more than a
decade (Dawes, 1994; Stivers, 1999).

Fourth, a reliable trend within the psychotherapy
outcome literature is that the
closer the study comes to real life clinical settings,
then the less significant the outcomes
tend to be (Epstein, 1996).

For instance, the
recent American multi-centre research trial
known as the `Fort Bragg demonstration
project' involved the analysis of the treatment
of 42,000 clients (who were largely
children) over a span of five years.

Yet the
results were disappointing in that there was
no evidence that psychological therapy led to
improvement in the lives of these recipients,
many of whom were struggling with significant
social adversities.

As the clinical psychologist
Tana Dineen (1999, p.128) observes :

these results should raise serious doubts about
some current clinical beliefs about the effectiveness
of psychological services . there is
scant evidence of its effectiveness in real life
settings.

What then can we conclude about the
effectiveness of psychotherapy?

Overall, the findings examined in this paper
highlight the overwhelming import of `nonspecific
effects' in psychological treatment
on the one hand, and of the frequently
flawed nature of RPCT methodology, on the
other.

None of which seems to be very encouraging
for the official view of psychological
therapy as a well-validated body of effective
clinical treatments.

Instead of engagement,
however, the tendency inside the main ther
apeutic professions seems to have been to
ignore or downplay these considerations
(see Howard, 2005), and it therefore seems
worth asking why this is so.

Personal conviction
is doubtless one of the reasons.

Within
the confines of the consulting room both
therapists and their clients will often observe
that the latter seem to undergo a significant
relief from their distress.

As already indicated,
this is one instance where immediate personal
experience can be compelling but also
highly deceptive, especially when backed up
by prevalent cultural myths.

Another element in this situation may be
the reliance of the field upon large-scale
meta-analytic studies, a trend that is reinforced
by the accumulating NICE guidelines
on psychotherapeutic practice in the NHS.

Although officially presented as both definitive
and authoritative (see for example,
NICE, 2003) such methods are notoriously
prone to generating misleading or inconclusive
data, as the previous discussion has
shown.

The meta-analytic approach simply
fails to capture the way in which knowledge
is developed and validated within the wider
scientific community.

When pursued in good
faith, scientific knowledge emerges from a
craft-like exploration and sifting of ideas
against the limits of personal experience and
of reasoned reflection - and not from managerial
directives or the behest of professional
interest groups, operating under the guise of
impersonal authority (Charlton, 2000; and
see Polanyi, 1955).

The significance of all this becomes clearer
when set against the evidence that - contrary
to the claims of Layard - our current social
arrangements may continue to underpin
much of the distress that brings people to
the consulting room (see, for example,
Perelman, 2005; Vail et al., 1999; Wilkinson,
1996).

In this situation, the interests of the
therapeutic professions are likely to dovetail
only too well with those of a political order
that is intent upon convincing its citizens
that their private troubles have little connection
with events in the public realm.

And in this respect, the myth of individual psychological
therapy as cure fits the bill admirably.


Affiliation

Adult Learning Disability Service, Telford.
Address

Adult Learning Disability Service, 40 Tan
Bank, Wellington, Telford, Shropshire
TF1 1HW

References

Assay, T.P. & Lambert, M.J. (1999). The empirical
case for common factors in therapy: Quantitative
findings. In M.A. Hubble, B.L. Duncan, & S.D.
Miller (Eds.) The heart and soul of change: What
works in therapy (pp. 33-56). Washington DC:
American Psychological Association.
Bergin, A. & Garfield, S. (1994). Handbook of psychotherapy
and behaviour change. New York: John
Wiley.
Bohart, A.C. (2000). The client is the most important
common factor: Clients' self-healing capacities and
psychotherapy. Journal of Psychotherapy Integration,
10(2), 127-149.
Boyle, M. (2002). Schizophrenia: A scientific delusion?
(second edition). London: Routledge.
Breggin, P. & Cohen, D. (1999). Your drug may be your
problem: How and why to stop taking psychiatric drugs.
Reading, MA: Perseus Books.
British Psychological Society (2002). Criteria for the
accreditation of postgraduate training programmes in
clinical psychology. Leicester: British Psychological
Society (Division of Clinical Psychology).
British Psychological Society (2004). Board of Examiners
in Counselling Psychology: Regulations and syllabus for
the qualification in counselling psychology. Leicester:
British Psychological Society.
Casement, P. (1995). Further learning from the patient:
The analytic space and process. London: Tavistock.
Charlton, B. G. (2000). The new management of scientific
knowledge: A change in direciton with
profound implications. In A. Miles, J.R. Hampton
& B. Hurwitz (Eds.) NICE, CHI and the NHS reforms:
Enabling excellence or imposing contro l ? London:
Aesculapius Medical Press (pp.13-32).
Cushman, P. (1995). Constructing the self, constructing
America. Cambridge, MA: Perseus Publishing.
Dawes, R. (1994). House of cards: Psychology and psychotherapy
built on myth. New York: Macmillan.
Department of Health (2001). Treatment choice in psychological
therapies and counselling. London: HMSO.
Dineen, T. (1999). Manufacturing victims: What the
psychology industry is doing to people. L o n d o n:
Constable.
Eisner, D. (2000). The death of psychotherapy: From Freud
to alien abductions. New York: Preager.
Epstein, W. (2006). The civil divine: Psychotherapy as
religion. Reno, NV: University of Nevada Press.
Epstein, W. (1996). The illusion of psychotherapy. New
York: Transaction.
Frank, J.D. & Frank J.B. (1991). Persuasion and healing:
A comparative study of psychotherapy. Baltimore,
MD: Johns Hopkins University Press.
Fonagy, T. & Roth, A. (1996) What works for whom? A
critical review of psychotherapy research. New York:
Guilford Press.
Furedi, F. (2003). Therapy culture: Cultivating vulnerability
in an uncertain age. London: Routledge.
Hansen, S., McHoul, A. & Rapley, M. (2003). Beyond
help: A consumer's guide to psychology. Ross-on-Wye:
PCCS Books.
Hawton, K., Salkovskis, P., Kirk, J. & Clarke, D. (1989).
Cognitive behaviour therapy for psychiatric problems: A
practical guide. Oxford University Press.
Holmes, J. (2002). `All you need is cognitive behaviour
therapy?' British Medical Journal, 324, 288-291.
Howard, A. (2005). Identity and counselling. London:
Palgrave.
Kelly, A.E. (2000). `A self-presentational view of psychotherapy'.
Psychological Bulletin, 126(4), 475-494.
Kline, P. (1988). Psychology exposed or The Emperor's new
clothes. London: Routledge.
Mahrer, A. (1998). `Embarrassing problems for the
field of Psychotherapy'. BPS Psychotherapy Section
Newsletter, 23(3), 3-25.
Mair, K. (1992). The myth of therapist expertise. In
W. Dryden & C. Feltham (Eds.) Psychotherapy and
its discontents. Buckingham: Open University.
Mcleod, J. (1994). Doing counselling research. London:
Sage.
Mearns, D. (1994). Developing person-centred counselling.
London: Sage.
National Institute for Clinical Excellence (2003).
D e p ression: The management of depression in primary
and secondary care (second consultation draft,
December). London: NICE.
Perleman, R. (2005). Manufacturing discontent: The trap
of individualism in corporate society. London: Pluto
Press.
Polanyi, M. (1955). Personal knowledge: Towards a postcritical
philosophy. London: Routledge.
Roth, T. & Stirling, P. (2006). Expanding the availability
of psychological therapy. Retrieved 12
February 2006 from www.bps.org.uk/dcp/news/
layard.cfm.
Smith, M.L., Glass, G.V & Miller, T.I. (1980). The benefits
of psychotherapy. Baltimore MD: Johns Hopkins
University.
Stivers, R. (1999). Technology as magic: The triumph of
the irrational. New York: Continuum.
Rosenthal, R. & Rubin, D.B. (1978). Interpersonal expectancy
effects: The first 345 studies. Behavioural
and Brain Sciences, 3, 377-415.
Tolman, C. (1994). Psychology, society and subjectivity: An
introduction to German critical psychology. London:
Routledge.
Vail, J. (1999). Insecure times: Living with insecurity in
contemporary society. London: Routledge.
Wilkinson, R. (1996). Unhealthy societies: The afflictions
of inequality. London: Routledge