ASK THE RIGHT QUESTIONS IN THE RIGHT WAY
Treatment research has been asking the wrong questions in the wrong way, resulting in “disappointingly negative results”, reveals a paper* by Jim Orford. We summarise the need for a “paradigm shift” in research on psychological treatments for addiction, and a willingness to confront the established culture.
Download AddictionToday130-Right research questions-Jim Orford
The dominant way of studying psychological treatments in the addiction field is faulty and has reached a dead end. In trying to gain knowledge of how the effectiveness of psychological treatments for addiction problems could be improved, researchers have looked in the wrong place in the wrong way. This article lists eight main faults and three key ways to address these – all identified by emeritus professor Jim Orford* – so that future research will help more clients into recovery.
1: EXISTING RESEARCH IGNORES ‘outcome equivalence paradox’ –
Different psychological treatments are often listed and scored in terms of the numbers of studies which appear to support their effectiveness. Some, mainly cognitive behavioural, have more research studies on them than others, including psychodynamic or counselling methods. A greater number of studies has been counted as greater effectiveness. But when different forms of treatment are compared directly in the same study, they are more likely to be shown as equivalent.
Also, comparing treatments can reflect the bias of researchers’ treatment allegiances.
2: EXISTING RESEARCH SACRIFICES RELATIONSHIP TO TECHNIQUE
The prevailing model of psychological treatment can be described as a technology model: supported by a manual and good training and supervision, it can be delivered to a high standard so that ‘therapist differences’ cease to be important. Voices against this say that the essence of psychological treatment is not the technique but the therapist-client relationship. In the UK Alcohol Treatment Trial, clients attributed positive changes mostly to this relationship. The evidence from Project Match, other addiction-treatment studies and psychotherapy generally is that, the better the ‘working alliance’ between client and therapist, the better the outcome.
3: EXISTING RESEARCH IGNORES research on unaided change
One of the great promises of the transtheoretical model of addictive behaviour change was that the stages and processes of change included in it would apply to both different treatments and to unaided change. In practice, the two fields of research are largely separate, which does not give us a complete picture. More research is needed.
Addiction Today note: future research should not merely describe participants as “heavy drinkers/users” but accurately diagnose whether they are addicted vs misusing, and for post-traumatic stress disorder and other dual diagnoses. Not only will this give a truer picture but we hypothesise that such people need professional treatment more than those without this diagnosis. Differentiating between the two should lead to more cost effectiveness and clinical effectiveness.
4: IT IMPOSES AN INAPPROPRIATE TIMESCALE ON CHANGE
The forms of treatment in trials such as Project Match and Ukatt are designed to be completed in weeks, with followup limited mainly to 12 months. High rates of relapse soon after treatment, and return to earlier stages of change, are known about but have not impacted the short-term focus of addiction-treatment research. A longer-term focus would be harder to manage within existing research and service funding, but there is growing support for it. This would be in keeping with growing interest in life-course trajectories of health more generally as well as addiction.
5: IT FAILS TO TAKE A SYSTEMS OR SOCIAL NETWORK VIEW
In the case of a treatment trial, an important procedure is the research assessment session(s) before treatment. The Project Match assessment was about six hours and the Ukatt one 2-3 hours – and at followup, some clients attributed change to the assessment.
Around the focal treatment there can also be other forms of treatment and assistance received in the same treatment agency. Even in a controlled trial, these cannot be completely eliminated.
The assistance can also come from elsewhere, including professional or mutual-help contacts.
And treatment clients continue to occupy the everyday world which consists of a myriad of encounters and events, any one of which could influence alcohol/drug use. Evidence shows that quality of home and job environments, and friendship and extended family resources and social support for change are related to positive outcomes after treatment or advice.
6: IT IGNORES THERAPISTS’ TACIT THEORIES
‘Local’ or ‘insider’ knowledge is not afforded a privileged place in conventional research. Researchers are often surprised that the theoretically-based treatments they devise and evaluate are not adopted by therapists in practice to the extent and with the fidelity for which the researchers hoped. That might indicate that therapists believe they are doing something which our research is not taking into account.
It would not be surprising, for example, if therapists’ own theories of action placed more weight on more complex kinds of matching, based on combinations of client variables and/or moment-to-moment or session-to-session matching of method to perception of a client’s needs as they arise during therapy.
7: IT EXCLUDES THE PATIENT’S VIEW
The addiction field could be accused of lagging behind newer ideas in the health services and sciences, where the involvement of service users in thinking about services, professionals sharing decision-making about the choice of treatment, and the promotion of partnerships with members of disadvantaged communities are valued, to ensure service appropriateness and accessibility.
At the same time, there have been moves to non-traditional research which aim to understand the views of patients in greater detail such as with qualitative research, and to conducting research giving participants a more active role in it, known as participatory research. We read of advocacy for mixing methods, including qualitative and quantitative.
However, journals often have a biomedical orientation, editorial policy and referees’ opinions which do not encourage non-traditional methods. They still hold a view that clients’ views on treatment and change are worth less than objective data. But clients are the chief protagonists and thus arguably the chief experts.
8: IT IGNORES DEVELOPMENTS IN SCIENCE PHILOSOPHY
The seven previous shortcomings fail to move with a trend of the last decades of the 20th century: critique of logical empiricism by philosophers of science – a recurring feature of which is that there are different forms of knowledge production.
In “mode 1” knowledge production, “logical empiricism” involves a researcher who must, as much as possible, be neutral and value free, testing hypotheses derived from scientific theory, using standardised ‘objective’ measures. Considered unimportant are a statement of the researcher’s views and experience, a description of the research setting, the meaning which participants ascribe to the assessment methods and interventions, and their experience of the research.
“Mode 2” assumes that there are important alternative forms of knowledge which are more local or ‘situated’, based on first-hand experience – for example the experience of being someone seeking treatment for addiction, or the experience of being a therapist for such a person.
Perhaps because so much of it is rooted in a biomedical tradition, the field of addiction treatment research remains on the whole resistant to a new philosophy of how to do science. That self-confidence is at odds with failure to advance understanding of addiction treatment and change.
SUGGESTED SHIFT 1: FOCUS ON STUDYING THE CHANGE PROCESS –
instead of studying named techniques. This is aimed mainly at dealing with failings 1-3. In the psychosocial treatment of addiction, there is growing support for the existence of important change processes common to treatments with different names and theoretical rationales.
Another promising line of treatment process research focuses attention on the relationship between clients and treatment staff. Assessment of the therapeutic alliance is among the options. Methodologically, this is quite hard, as it cannot be assessed simply at the outset of treatment or at followup, but is an emergent property of the interaction between two or more people.
A complementary approach could focus on events during treatment, similar to the “events paradigm”, such as clients’ use of commitment-to-change statements, completing homework assignments and recommended family activities.
So one way forward is to develop and test a theory, or better theories, of addiction behaviour change. To move in the direction advocated here, such a theory would need to specify therapist actions, client actions and features of the developing client-therapist relationship.
The theory might draw on a model of processes of change that transcends treatment types: for example, self-liberation or consciousness raising, or client-therapist consensus about the aims of therapy, or therapists’ and clients’ belief in the ability of their treatment to effect change.
SUGGESTED SHIFT 2: STUDY THE CHANGE PROCESS WITHIN BROADER, LONGER-ACTING SYSTEMS
Failings 4-5 might require a more radical break with tradition. We might start by studying the ‘therapeutic climate’ of the organisation setting in which treatment occurs. Moos and his colleagues showed that positive outcomes are related not only to the key therapeutic relationship but also to the perceived high quality of relationships with the whole treatment team, high expectations for personal growth and change engendered by the treatment environment and moderate organisational structure. We should examine all client procedures in our search for sources of change, including referral and treatment-entry procedures. We might have to seek collaborators with knowledge of wider systems.
Our theories of change, and our research methods, should broaden to include settings and systems that we already know to be important but which our theories of change and our institutional and disciplinary affiliations and allegiances keep at arm’s length. Candidates include Alcoholics Anonymous, Gamblers Anonymous and other addiction mutual-help organisations. They might include faith communities. We could also think of systems in the language of social networks.
SUGGESTED SHIFT 3: ACKNOWLEDGE THE VARIETY OF SOURCES OF USEFUL KNOWLEDGE
The third shift addresses the very nature of the science we practice and aims to deal with failings 6-8. In general, the call is for research which uses quantitative research less and qualitative research more, and which in its design involves closer collaboration between researchers and participants, drawing on traditions such as those of action and participatory research. This varied collection offers a greater ‘voice’ to participants.
As Sullivan said: “outcomes research is forcing us to recognise that only the patient can determine if medical treatment has been successful”.
*ORIGINAL PAPER AND REFERENCES.
Orford's peer-reviewed 11-page research paper, with full research references, was published by Addiction.
JIM ORFORD is emeritus professor at the Alcohol, Drugs, Gambling and Addiction Research Group at the School of Psychology, University of Birmingham.