Leichsenring sets out to describe research related to psychotherapy effectiveness, and in particular, for which psychiatric disorders randomized controlled trials of specific models of psychodynamic psychotherapy are available

Is Psychotherapy Any Good?
A Review of evidence relating to Psychodynamic Psychotherapy and the
Nature of Psychodynamic Assessment

Sean Manning
October 2006

Introduction
This document is intended as a resource to stimulate discussion around the
validity and effectiveness of psychodynamic psychotherapy and psychodynamic assessment. It is written in an atmosphere of, to this writer, unhelpfully argumentative debate between ‘scientific’ (symptom and behaviour oriented) clinical psychology, and the more descriptive, relationship-oriented psychotherapies, in which the latter have been criticized as unscientific and lacking evidence for their practice (Surgenor 2006). In 1952, Hans Eysenck published his famous critique, claiming that psychotherapy was no better than no treatment at all, which led someone to quip, ‘Psychotherapy is an undefined technique applied to unspecified problems with unpredictable outcomes.’ It would seem that his conclusion is still current among many ‘scientific’ practitioners, even after decades of research that demonstrate considerable evidence for the efficacy of psychotherapy. Despite this controversy, or maybe because of it, many psychotherapists eschew scientific investigation. Psychodynamic therapists, and in particular psychoanalysts, often seem to regard the uninitiated with as much disdain as behavioural researchers have towards them for their allegedly poor science (Fonagy 2006). One consequence of this lack of co-operation between disciplines has been a poor showing of research into psychodynamic methods, compared to behavioural, symptom- oriented treatments. Thus behavioural therapists can claim to have more scientific basis for their work. This document is a summary of summaries, using several current reviews, a few more general resources and a couple of specific articles. It is not intended to be a rigorous scientific paper, but it is, I believe, a truthful reflection of what 1 The terms ‘psychodynamic psychotherapy’ and ‘psychotherapy’ are used interchangeably. The terms ‘psychoanalysis’ ‘analysis’ and ‘analytic therapy’ are also interchangeable, but indicate a particular form or forms of psychotherapy characterized by a reflective, interpretive and neutral therapist stance, and relatively long duration. This is the widely accepted definition of psychodynamic psychotherapy given by Gabbard; ‘a therapy that involved careful attention to the therapist-patient interaction, with thoughtfully timed interpretation of transference and resistance embedded in a sophisticated appreciation of the therapist’s contribution to the two-person field’. is ‘out there’ at present. It makes no exaggerated claims for psychodynamic psychotherapy. It has three parts. In the first part, the bulk of the paper, I will discuss some of the arguments and issues relating to research into psychotherapy. Doubts about the ‘gold standard’ of outcome research, the ‘randomized controlled trial’, it’s applicability to psychotherapy and the American Psychological Association’s list of ESTs – empirically supported therapies are discussed. Then such research as exists is summarised. Psychotherapy effectiveness in relation to specific disorders is reviewed, as well as evidence for long term psychotherapy. Part two deals with a different order of evidence. This is a brief discussion of the implicit direction indicated by the related literatures on attachment and narrative. Finally in part three there is a discussion of psychodynamic assessment, as distinct from clinical diagnosis, and a short discussion of the problems with PART 1: Research and Research Evidence

Problems with outcome research – 1 - the APA list of empirically supported
treatments

The ‘gold standard’ in outcome research is the randomized controlled trial
(RCT), in which subjects are randomly allocated to one of two or more
conditions - ‘experimental’ or ‘control’. The standard used by the American Psychological Association’s (APA) Task Force on Promotion and Dissemination of Psychological Procedures for ‘Level 1’ (the best) evidence is two or more randomized controlled trials covering the same ground (Leichsenring 2005). In 1995 the APA task force began to report on ‘empirically supported treatments’ - ESTs - and a list was compiled. This list, regularly updated, still defines the international standard for psychological treatments. The case for the list is that it empowers consumers, giving some protection from fringe therapies; provides information about what works so that everyone can make better decisions, including funders; and that it improves the training and education of practitioners. The case against, as summarized in a recent Scientific American – Mind article (Arkowitz and Lilienfeld 2006) gains momentum with time. ­ The list is not verified in the real world. As Glen Gabbard said in a 2004 New York Times interview - 'The move to worship at the alter of these scientific treatments has been destructive to clients in practice, because the methods tell you very little about how to read the real and complex people who actually come in for therapy.' Too much is sacrificed - therapist flexibility is constrained, researchers reject up to 90% of subjects in the interests of homogeneity and the all-or-nothing nature of the list omits information, like measures of degrees of efficacy. Some treatments have quite modest effects, with many subjects not helped at all, or likely to relapse. Westen and Novotny (2001) showed how the figure of 51% of depressed patients who improved following CBT dropped to 37% 1-2 years later. ­ The EST list is biased in favour of cognitive-behavioural techniques. Reviews of psychoanalytic and humanistic treatments suggest equally positive effects, but there is less research on them, so less likelihood of being included. ­ ESTs focus on symptoms and distress to the exclusion of other factors ­ The technique may not be what produces the change. The 'Dodo-Bird verdict' – confounds researchers. In Alice in Wonderland, the Dodo Bird, judging a race, announced, 'Everybody has won and all must have prizes'. There is a tendency for all psychological or psychotherapeutic therapies to do equally well. In response to these arguments and anomalies, in 2005 the APA issued a policy statement defining evidence-based practice as 'the integration of the best available research with clinical expertise in the context of patient
characteristics, culture, and preferences.' This amounts to recognition that
regarding randomized controlled trials as the gold standard might have been an
error.
Problems with outcome research 2
The difficulty with randomized controlled trials - RCTs

Randomization in research serves to eliminate differences between groups,
thus controlling potentially disturbing influences on differences caused by
treatment. Randomization is thought to ensure the internal validity of a study. In terms of believable evidence, this is certainly the case. However, there are a series of problems in applying RCTs to psychotherapy methods. Research into therapy that is considered empirically supported tends to have three characteristics. 2 Quoted in the Scientific American Mind article. 1. Studies address a single disorder, usually Axis I of the DSM-IV. The use of Axis I as a basis for research on treatment effectiveness has the advantage that the subject of inquiry can be treated as a discrete disorder with a measurable set of symptoms, thus producing apparently clear results for ‘evidence-based’ treatment. The disadvantage is that the co-morbidity commonly observed in clinical practice, often involving Axis II diagnoses (personality factors) which so often affects treatment outcome, is ignored. 2. Treatments are manualized and are of brief and fixed duration, normally without flexibility related to the individual patient. Psychological disorders are treated rather like physical illness. 3. Outcome assessment is based on change in the target symptom. There is a trade-off between the undoubted internal validity of controlled trials, where the sample is deliberately homogeneous and often small to minimize the effects of extraneous factors that may affect treatment, and the external validity of ‘naturalistic’ research on (usually) larger populations of real patients presenting at treatment facilities. In the former case, the relationship between the treatment being studied and outcome is clearer, but the situation is artificial and less generalisable to everyday practice. In the latter, results can be more confidently generalized, but there are issues with co-morbidity (more than one diagnosis diagnoses) and heterogeneity in the sample. Westen et al. (quoted in Fonagy, 2006:766) have identified four poorly supported assumptions underpinning the use of this kind of outcome research: 1. psychopathology is so malleable that brief intervention can change it 2. most patients can be treated for a single disorder; 3. psychopathology can be treated with psychosocial interventions without regard to personality factors that are less likely to be responsive to brief 4. randomized controlled trials are the best way to evaluate treatment In reality, most forms of psychopathology encountered in clinical practice are treatment resistant, co-morbid with other disorders, and exist in the context of a distinct and idiosyncratic personality structure. A series of problems arise when the randomized research design is applied to a humanistic field like psychotherapy. Treatments which rely for their evidence on RCTs tend to place the emphasis on the disorder rather than the person. No matter how useful the comparison for science and medicine, people, especially in their psychological functioning, are not like machines. This introduces a series of difficulties related to the categorical diagnostic approach, such as is used in the Diagnostic and Statistical Manual (DSM) and the International Classification of Diseases (ICD), the two main international standards. This is discussed further below, under ‘Assessment’. For now, let us suggest that the psychological functioning of people is best assessed along a series of dimensions, or continua. For instance, a person may be more or less anxious, impulsive, depressed or content, rather than categorically suffering from an anxiety disorder, a depressive disorder or an impulse control deficit. Because RCTs generally focus on diagnostic categories, the results often miss the idiosyncrasies of the person. 2. Artificial conditions It is questionable whether RCTs are representative of clinical practice (Leichsenring 2005). Randomized trials tell us only one thing – that a treatment is or is not effective under the controlled conditions of the study. The trial condition often bears little resemblance to real clinical practice. If a method of psychotherapy has been shown to work under the controlled conditions of the RCT, it is assumed, based on the medical disease model, that it will work under any conditions. A drug may be assumed to work equally well under a variety of conditions, but psychotherapy is not a drug and the disease model is not adequate to describe complex psychological difficulties. 3. Co-morbidity In normal psychotherapy practice, complex conditions with a high degree of co-morbidity are encountered. That is, a typical client may not easily fall into a single diagnostic category – they may satisfy the criteria for several diagnoses, or have co-occurring features of several, without actually meeting the conditions for any. All of these features may receive attention from the
clinician at the same time. This is not typical of research conditions.

4. Manualized treatments

In an RCT, a manual or structured method will be followed as closely as
possible. In real practice, manuals are tools to be used when the clinician
thinks they are appropriate, often in conjunction with a series of techniques
and interventions that are not in the manual. The production of manualized therapy, a typical outcome of RCT-style research, may have a long-term effect on clinical practice by narrowing and damaging the breadth of clinical training. A well-trained clinician, even when using a standardized approach, will deviate from it according to the responses of individual patients, and according to the clinicians responses to the patient. This flexibility and empathic responsiveness is limited in graduates who have been trained to follow the manual, and who have not had training in personal reflection and empathic responsiveness. 5. Applicability to long term treatments RCTs are not applicable to long-term therapy (lasting several years) or to psychoanalytic therapy. It is not possible to carry out a manualized therapy for several years. Inevitably the therapist will stray from the strict terms of the method. It is not possible or ethical to assign patients to a “no treatment” or “treatment as usual” group for several years if a special treatment is available. Patients who opt for longer-term therapy differ in personality traits from those who choose a shorter method. Random assignment of patients would destroy the patient-therapist matching and invalidate the therapeutic approach. The results of a randomized trial for longer-term psychoanalytic therapy would not therefore be valid for the patients who normally seek such therapy (Leichsenring 2005). 6. Some people resist randomization Bateman and Fonagy (2004) suggest a number of problems with outcome research, pointing out that borderline clients do not take kindly to randomization. “Although their lives may be dominated by apparently random behaviour, their search is for stability, certainty, control. … when they realize that their allocation to treatment appears to be dependent on the toss of a coin they are confronted with uncertainty, loss of control and anxiety about rejection. … the result may be, at best, a feeling of demoralization, and, at worst, rage and aggression and refusal to participate…” (Op. cit. p 52) This may be a reason for observed high attrition rates in control groups – that is, randomization is iatrogenic. Most studies of borderline patients compare a special treatment with ‘treatment as usual’ (TAU). One thing that such patients do not like is to be considered routine, especially if it is known that there is another treatment available. (Op. cit. p 53) There is a literature on expectations about therapy. Trials that are incongruent with expectations are likely to produce poor outcomes. 8. Attrition Attrition represents a serious threat to internal validity in typically small-sample controlled trials. 9. Therapist randomization With few exceptions, therapists are not randomized to patients, despite evidence that the personality, skills and training of the therapist have a considerable impact on outcome. Comparisons are often made between psychodynamic psychotherapy and cognitive methods. The latter can certainly lay claim to ‘evidence-based-practice’, if only for the reasons set out above, that psychodynamic therapies are less amenable to randomised controlled trials. Jeremy Holmes (quoted in Fonagy, 2006) has argued convincingly in a 2002 edition of the British Medical Journal that: 1. the foundations of cognitive therapy are less secure than has been 2. the impact of CBT on the long term course of psychiatric illness is not 3. in at least one ‘real time’ trial, CBT had to be discontinued because of 4. the effect size of CBT is exaggerated by comparison with wait-list 5. there is evidence of ‘post-cognitive’ behaviour therapy approaches that increasingly take psychodynamic concepts into account. In response to these arguments it has been suggested (Nick Tarrier, also BMJ 2002, quoted in Fonagy 2006) that; 3 ‘Effect size’ is the difference in mean outcomes of the treatment and control group divided by the standard deviation of the outcomes of the control group. Essentially it is expressed as a percentage of one standard deviation. An effect size of 0.7 is substantial. 1.0 or above is exceptional. ‘… [traditional psychotherapy] has been around for 100 years or so. The argument … becomes a little less compelling when psychotherapy’s late arrival at the table of science has been triggered by a threat to pull the plug on public funding because of the absence of evidence.’ Psychodynamic clinicians are invited with some justification to; ‘… do more than gripe and join in the general endeavor to acquire data.’ It is in this spirit that the following review if offered.

Review of evidence

A general conclusion of the Task Force of the APA on Promotion and
Dissemination of Psychological Procedures is that psychotherapy is ‘probably
efficacious’. The caution in this statement reflects the difficulty in finding
randomized trials of psychotherapy, particularly finding two studies that
replicate results.
A similarly general conclusion was reached by Gabbard, Lazar et al. (1997) (reviewed by Leichsenring) who conclude that psychotherapy is a cost-effective treatment when compared with other health costs. Three meta-analyses have been used here to briefly summarise research into psychotherapy effectiveness. Leichsenring (2005) sets out to describe research into psychotherapy effectiveness, and in particular to discover for which psychiatric disorders randomized controlled trials of specific models of psychodynamic psychotherapy are available. Considering the difficulties of using randomization with psychodynamic treatments, Fonagy and others, in two major reviews (Fonagy, Roth et al. 2005; Fonagy 2006) have set out to discover: 1. are there any disorders for which short term (20 sessions or less) psychodynamic psychotherapy can be considered evidence-based; 2. are there any disorders for which psychotherapy is uniquely effective as either the only evidence-based treatment or as a treatment that is more effective than alternatives, and 3. is there any evidence base for long-term psychodynamic psychotherapy? Leichsenring details randomized trials available for: 4 Most of this work refers to short term psychodynamic psychotherapy (STPP), meaning 20 sessions or less (USA) or 50 sessions or less (Germany). Anxiety disorders (1 trial) Post-traumatic stress disorder (1 trial) Somatoform disorders (4 trials) Bulimia nervosa (3 trials) Anorexia nervosa (2 trials) Borderline personality disorder (2 trials) Cluster C personality disorder (1trial) Substance-related disorders (4 trials) Aside from RCTs, ‘process research’ - looking at what happens in therapy - and ‘effectiveness research’ - looking at outcomes in real clinical practice, for instance in terms of health spending and insurance company data, are also surveyed. Depression According to Fonagy about 20 short-term psychodynamic trials have been published in relation to depressive or anxiety disorders or symptoms. In general it has been shown to be more effective than wait-list or outpatient ‘treatment as usual’. In Fonagy’s review comparison with cognitive behaviour therapy tends to come down on the side of the behavioural methods. However, in a number of cases, therapies in these trials considered ‘psychodynamic’ were not actually bona fide therapies. This reveals researcher bias expressed in a lack of investment in the ‘control’ condition. Where there are better designed and implemented alternatives to behavioural treatment, the differences tend to disappear. There tends to be no superiority for CBT on follow-up, and differences in effect size are limited with severely depressed patients. It still seems likely that symptom-oriented behavioural treatment might produce better results in the short term, with this difference disappearing or even being reversed in the longer term. Typical of such results is the ‘Sheffield Psychotherapy Project’ where it was found that behaviour treatment was superior to psychodynamic therapy after 8 sessions, but not so after 16. The Helsinki Psychotherapy Study suggests similarly that short-term symptomatic remission is better with a behavioural treatment – solution focused therapy – than short term psychodynamic psychotherapy, but personality disorder factors were more affected by the latter. The inclusion of interventions specific to a psychodynamic approach is associated with better outcomes even in therapies born of a different Leichsenring finds four randomized trials in which psychotherapy and cognitive behavioural methods are found to be equally effective, with large effect sizes that are stable over time. In combination with pharmacotherapy, psychotherapy improves outcome over either drug treatment or psychotherapy alone. Nevertheless, evidence for psychodynamic therapy in depression is weak – there are few compelling demonstrations, and none for long-term approaches. As Fonagy puts it; ‘… the data are consistent with the assumption that a proportion of patients in any research sample will respond to therapeutic intervention of any kind.’ Anxiety Included here are phobia, generalized anxiety disorder, panic disorder, post-traumatic stress disorder and obsessive-compulsive disorder. This field is dominated by cognitive-behavioural packages, compared with often poorly-structured control ‘therapies’. Again, where the ‘placebo’ is an active treatment, the effect of behavioural treatments is diminished. However, the absence of studies with a well-structured psychodynamic alternative makes it difficult to state anything definitively. Where complicated grief and bereavement reactions (as opposed to exposure to trauma) are involved, psychotherapy may be more effective (Fonagy 2006). In the single trial reviewed by Leichsenring, moderate-length psychotherapy of 36 sessions proved as effective as behavioural methods for generalized social phobia. Post-traumatic Stress Disorders A single trial reviewed by both Leichsenring and Fonagy concludes that psychotherapy is as effective as trauma de-sensitization. Trials reviewed by Fonagy suggest positive results for psychotherapeutic approaches that investigate the meaning of the traumatic event for the person’s sense of self and place in the world. However, methodology is problematic, and results cannot be confidently quoted. Somatoform Disorders The three trials reviewed by Leichsenring suggest that psychotherapy is more effective with about two-thirds of patients than standard medical treatment and as effective (with irritable bowel syndrome) as treatment with paroxetine.
Psychotherapy (but not paroxetine) was associated with a significant reduction
in healthcare costs over time.

Eating Disorders

Four available trials in this area suggest that psychotherapy is as effective as
other treatments. Where it is modified for a specific clinical problem, it does
as well as behavioural treatments which are similarly focused. The exception
is bulimia, where behavioural treatment seems slightly more effective.
Substance Misuse In the short term, psychotherapy seems to have little impact on this group of disorders, where motivational enhancement and community support are the treatments of choice. As with other disorders, in the longer term, psychotherapy may produce better maintenance of gains. Its place is probably as a secondary treatment, after acute problems have been overcome, and then only for some clients. Personality Disorder There has been more research attention paid to this diagnostic group, allowing more than one meta-analysis. Substantial effect sizes of 0.7 – 1.3 have been found for psychodynamic psychotherapy, and a comparison with cognitive-behavioural treatment reveals effect sizes of 1.31 for psychotherapy, based on 8 studies, as opposed to 0.95 for cognitive behaviour therapy, based on 4 studies (Fonagy 2006). Broadly, the approaches seem equally effective. Bateman and Fonagy (2004) review evidence for treatments in relation to borderline personality disorder. Treatments shown to be effective (p 44) tend: ­ to be well structured; ­ to devote considerable effort to enhancing compliance (building a ­ to be clearly focused, for instance on problem behaviour such as self- harm or on problematic relationship patterns; ­ to be theoretically coherent to both therapist and patient, sometimes deliberately omitting information incompatible with the theory; ­ to be relatively long term – for example 2 years or more; ­ to encourage a powerful attachment relationship between therapist and patient, enabling the therapist to adopt a relatively active rather than passive stance, and ­ to be well-integrated with other services available to the patient. When compared with a psychiatric treatment-as-usual group, patients with BPD treated with a psychotherapeutic approach involving partial (in-patient and day) hospitalization tend to improve on a variety of measures, including; ­ frequency of suicide attempts and self-harm, ­ number and duration of in-patient admissions, ­ use of psychotropic medication, ­ self-report measures of depression, anxiety, general symptom distress, interpersonal function, and social adjustment. In another study, also quoted by Bateman and Fonagy, not only were these gains maintained, but patients continued to improve on most measures after the termination of treatment. Health service costs were the same as the TAU group before and during treatment, but significantly less on follow-up. There is evidence supporting cognitive and ‘cognitive analytic’ treatments in relation to BPD, but early gains tend not to be maintained, and the natural instability of the BPD diagnosis can distort results that use it as a measure of outcome. (Naturalistic studies, looking at what happens in normal clinical practice, suggest that people diagnosed with BPD will not meet the criteria for the diagnosis for at least two consecutive months within the first 12 months of diagnosis.) (Bateman and Fonagy, 2004: 46) DBT (dialectic behaviour therapy), a special adaptation of CBT combined with meditation and ‘mindfulness’ – secular derivatives from Buddhism – designed specifically for people with BPD appears to have better retention rates than TAU and to reduce episodes of self harm in the short term, but seems less effective in the longer term. It may be effective in relation to certain behaviours, but not as a treatment for the personality disorder itself. More effective may be Bateman and Fonagy’s Mentalization-Based Therapy (MBT) (sometimes referred to as “understanding misunderstanding’) which concentrates on integration of the client’s thinking, feeling and intention in the moment. Transference is considered, but only as a tool to engage ‘mentalization’. (Bateman and Fonagy 2004). Psychotherapy appears effective as a part of therapeutic community (TC) treatment for Cluster B personality disorders and substance misuse. No methodologically sound trials of psychotherapy with antisocial personality disorder are quoted in the reviews summarized here. There is some suggestion that psychotherapy may produce gains in combination with TC treatment, but research methods are too weak for confidence. With cluster C personality disorders (anxious-avoidant), there is some evidence that psychotherapy is equivalent to cognitive therapy and possibly slightly superior. Drop-out rates are high with avoidant symptoms (sic) though better Long-term psychotherapy

There are few quotable studies for long term psychotherapy or psychoanalysis,
and these are not associated with specific disorders or symptom clusters.
Those that are available tend to be naturalistic, though within the constraints on this methodology, generally positive. Leichsenring is more positive than Fonagy about the validity of this methodology, concluding that naturalistic effectiveness studies, when compared with studies using randomization, do not exaggerate the effectiveness of short term psychotherapy, so are acceptable in relation to long term work. For long-term psychoanalytic therapy, effect sizes greater than 0.8 and up to 1.55 are found in relation to symptoms, interpersonal problems, social adjustment, in-patient days and other outcome criteria. There is some quite convincing evidence drawn from insurance company data, where health costs are crucial (effect size 0.78). There is some evidence that the effect continues to increase after termination of therapy. (Leichsenring, op. cit.) Psychoanalytic therapy appears to be particularly effective in relation to changes in personality, and may be more effective than shorter forms of psychotherapy for people with complex psychiatric disorders. There are some challenging results, however. For instance, those therapists whose attitude most closely resembles the classical psychoanalytic stance (neutrality, focus on insight) tend to get the poorest results. (Fonagy, 2006) A major German study of over 400 patients in analysis for a mean 6.5 years showed considerable stable positive changes in 70%-80% of the sample, with associated savings in health costs, but there are no pre-treatment measures, so effect size is impossible to estimate. (Leichsenring) There is some evidence that children with severe emotional disorders do quite well in analysis. (Leichsenring) In general, evidence on psychoanalysis favours those with milder disorders, though there are positive results for more severe disturbances. Process research suggests that outcome in psychodynamic psychotherapy is related to competent delivery of technique and the development of a therapeutic alliance. Fonagy (op. cit. p 806) quotes Whittle suggesting that a likely cause of the paucity of research into the effects of psychoanalysis lies in ‘fundamental incompatibilities in the world view espoused by psychoanalysis and most of current science.’ Tentative conclusions based on reviewing the research evidence
From this, we might begin to draw some tentative conclusions. Psychodynamic
psychotherapy seems likely to be effective with depression, anxiety,
somatoform disorders, eating disorders, Cluster C and possibly Cluster B personality disorders, and possibly as a secondary treatment with substance misuse disorders. It is likely that, where behavioural methods show initial superiority, this becomes less so with time. That is, psychotherapy takes longer to work. However, it is possible that its effects are longer-lasting and it may be more effective in the very long term at reducing health spending. Even more tentatively, we might suggest that the benefits of long term psychotherapy and psychoanalysis continue to improve after treatment is finished. There is reason for caution, however. While there is considerable evidence for both behavioural (usually justified by small randomized controlled trials) and psychodynamic (more often from naturalistic studies) treatments; ‘… there remains a long way to go before we can be sure about which patients are best suited to psychoanalytic, behavioural and supportive therapy, and who should be treated within what context and with what intensity.’ (Bateman and Fonagy 2004) Also mentioned in the Arkowitz and Lilienfeld article in Scientific American – Mind are some therapies to avoid. They include ‘energy’ therapies, in which a client’s ‘energy fields’ are manipulated by the therapist; recovered-memory techniques; rebirthing; ‘facilitated communication’ – a technique with autistic children, and crisis de-briefing – there is mounting evidence that it makes PTSD symptoms worse. PART 2: Evidence on Attachment Theory and Coherent Narrative that
Supports Psychodynamic Psychotherapy
There is a strong body of evidence linking attachment style to personal functioning over a wide range of measures, including personal comfort, material success, stability and satisfaction in personal relationships. Beginning with the study of children, the extension of attachment research into adult patterns of relatedness has had profound implications for the practice of psychotherapy and other psychological treatments (Schore 1999; Siegel 1999; Fonagy 2001; Fonagy, Gergely et al. 2002). This work considers four kinds of representational systems (Bateman and Fonagy 2004: 57): 1. expectations of early caregivers created in the first year of life and subsequently elaborated into relational expectations; 2. event representations by which general and specific memories of attachment-related experiences are encoded and retrieved (a kind of ‘filter’ on experience and memory of relationships); 3. autobiographical memories by which specific events are conceptually connected because of their relation to a continuing personal narrative and developing self-understanding, and 4. understanding of the psychological characteristics of other people – inferring and attributing intention, motivation, emotion and beliefs in others, and differentiating these from those in ourselves. Together, these constitute what we know as the ‘internal working model’, a template for our experience of our selves and others, closely related to the development and coherence of a ‘sense of self’ (Stern 1985). This work not only gives us models for understanding psychological distress and pathology, it also defines healthy functioning in terms of concepts like the ‘sense of self’ (Stern 1985), the ‘coherent autobiographical narrative’ (Siegel 1999), ‘mentalization’ and ‘reflective functioning’ (Fonagy, Gergely et al. 2002), ‘metacognitive monitoring’ (Main and Hesse 2001) and so on. Parents with a well-developed ability to reflect on their own functioning and that of others tend to pass on a secure attachment base to their children. On the other hand there is a close relationship between intergenerational trauma and abuse and a parent’s capacity, or lack of it, to reflect on her own and others’ thoughts and feelings, and on her own history (Bateman and Fonagy 2004: 75-79). Almost every measure of success and difficulty in life appears to be connected with the nature of a person’s attachment pattern. This work is well supported by studies of neurological development (Damasio 1999; Schore 1999; Siegel 1999; Cozolino 2002; Damasio 2003). The whole body of theory and research is supportive of the fundamental psychodynamic proposal - that there are basic processes underlying the way we function and how we produce symptoms which are best addressed in a relationally-based psychotherapy. This is not to denigrate the kind of symptom-based treatment such as is commonly provided by evidence-based practice in clinical psychology, but it is a serious error to suggest that the latter is a sufficient approach to human psychological discomfort. Not everything can be diagnosed from the history. Much of what becomes focal for therapy will emerge as the client engages with the therapist. Idiosyncratic attachment style will emerge in the dynamics of what we term transference and counter-transference. A relational therapy is essential to allow this to happen. Since interpersonal functioning is fundamental to human contentment and distress and to success on almost any dimension, it is necessary to have a form of therapy in which interpersonal elements are central. This leads to a discussion of psychodynamic assessment, below. No matter what the symptom, it is important to understand who has it and what meaning they make of it, as opposed to diagnosing a condition by listing symptoms and producing a treatment out of a manual. The ‘Dodo-bird verdict’ mentioned above makes sense when view from this perspective. PART 3: Psychodynamic Assessment
It has been mentioned above that the assessment of psychological functioning according to diagnostic categories presents some difficulties. Diagnostic categories can be used in outcome research to look at the effects of treatment on specific ‘disorders’, and can assist an apparently orderly process from assessment through diagnosis to treatment, much like in general medicine. This essentially mechanistic approach supposes that the human mind works in a similar way to body tissues, and that the whole resembles a piece of machinery. The human mind, however, has few categorical disorders – developmental disorders like intellectual disabilities, autism and possibly attention deficit, may with some justification be treated as discrete ‘illnesses’. Psychotic conditions sometimes behave like this, though they rarely present as clearly as their descriptions in DSM-IV and ICD-10 suggest. In general, psychological problems fit uncomfortably into diagnostic boxes. For instance, of the 9 criteria in DSM-IV-TR for borderline personality disorder, 5 are needed to make the diagnosis. Thus two people with this diagnosis need share only one symptom, and in other respects be quite different from each other. Auditory hallucinations – ‘hearing voices’ – especially when accompanied by fixed, false beliefs called ‘delusions’, along with ‘inappropriate’ or ‘flattened’ emotional states, is usually considered indicative of psychosis. Yet all of these symptoms can be caused by severe, repeated early trauma (Read 2005). A diagnosis of depression gives no indication as to the way a particular person experiences it, within their distinct personality structure. A difficulty recognizing thinking and feeling patterns in others might affect many aspects of life, might be expressed in a range of symptoms and signs, and may itself be a source of anxiety, yet a diagnosis of anxiety disorder would be misleading. As the introduction to the 2006 Psychodynamic Diagnostic Manual (PDM) observes; ‘Mental health comprises more than simply the absence of symptoms. It involves a person’s overall mental functioning, including relationships; emotional depth, range and regulation; coping capacities; and self-observing abilities. . Any attempt to describe and classify deficiencies in mental health must therefore take into account limitations or deficits in many different mental capacities, including ones that are not necessarily the source of pain.’ The oversimplification of mental health phenomena by reducing our problems to a series of check lists serves us ill. Reliability among clinicians is poor (PDM, p 3) compared to physical medicine, and understanding of essential developmental and self-regulating processes is not assisted by listing clusters of symptoms. Co-morbidity may be usefully considered an expression of underlying processes, but this is confused by using categorical diagnosis. Oversimplification in the interests of measurement is bad science. For instance, Karlen Lyons-Ruth (1995), studying parental depression, remarks that categorical diagnostic thinking tends to ascribe many difficulties, for instance, marital conflict, to the depression, while interpersonal difficulties are often observed to persist after the depressive episode is over. She notes that; ‘Mood rarely occurs in a vacuum but is intricately linked to the flux of interpersonal successes and failures in life, as mediated by one’s implicit representational systems for organized meaning and values and by the economic resources available to realize those life aims.’ Lyons-Ruth, pleading for a reintroduction of implicit representational systems into psychotherapy, suggests that these, along with associated interpersonal strategies (attachment patterns), and associated dysphoric affect are less episodic and more stable, individually and intergenerationally, than is commonly supposed. In one study, relationship patterns in depressed mothers’ families of origin accounted for all of the variance in parenting behaviour associated with depression. In other words, long-term implicit relational patterns account for the problem better than the ‘illness’. In a relational-systems model (as opposed, in this case, to a mood-disorder model) ‘the type and severity of relational dysfunction … would become as important as the type of symptom expression.’ What is required is a way of describing not just symptoms, but internal experience and a full range of mental functioning. This is the task of psychodynamic assessment. A good example of assessment is provided by the PDM, a co-operative project of a number of psychoanalytic organisations. It uses a multi dimensional approach, with three axes. The first (the ‘P Axis’) tells us about personality structure, both in terms of type and degree of health or pathology, because; ‘… symptoms or problems cannot be understood, assessed or treated in the absence of an understanding of the mental life of the person who has the symptoms.’ (PDM p 80) The second (the ‘M Axis’) provides detail of the person’s mental capacities – regulation, intimacy, quality of experience, affective experience, defensive patterns, internal representations, integration, self-observation and standards The third (the ‘S Axis’) describes symptoms and symptom clusters, emphasizing the person’s subjective experience, rather than considering symptoms to constitute categorical phenomena. This approach, highly descriptive, very personal, taking into account the experience and uniqueness of the individual, compliments the established diagnostic approaches. Its scientific basis is documented, and the rationale for each of the elements appears to be sound.
Final Comment

The making of meaning is central to basic functional processes. Our ability to
do this is related to our attachment experience, and the whole attachment/
meaning-making enterprise is fundamental to human nature (Fonagy 2006). There has been economic and political pressure from short term, cheaper, symptom-based psychological treatments and drug regimes. However, it is not reasonable that some process resembling psychodynamic psychotherapy or psychoanalysis with its emphasis on relationship, attachment and meaning could ever be considered inappropriate among the range of approaches available to people with mental health problems. There always has to be a concern with the psyche as well as a monitoring of symptomatology. To turn away from the essential human (perhaps mammalian) quality that allows us to recognize in ourselves and others the presence of a mind, of desires, emotions, intentions and beliefs; ‘. risks apocalyptic cultural and social changes . if we mock personal experience and start to deride the feelings, thoughts, and desires of our fellow humans.’ (Fonagy 2006:774) References

Arkowitz, H. and S. O. Lilienfeld (2006). Psychotherapy on Trial. Scientific
American Mind. 17: 42-49.
Bateman, A. and P. Fonagy (2004). Psychotherapy for borderline personality disorder - mentalizaton-based treatment. Oxford, OUP. Cozolino, L. (2002). The Neuroscience of Psychotherapy. NY, London, Norton. Damasio, A. (1999). The Feeling of What Happens. London, Randon House. Damasio, A. (2003). Looking for Spinoza - Joy, Sorrow and the Feeling Brain, Harcourt. Fonagy, P. (2001). Attachment Theory and Psychoanalysis. NY, Other Press. Fonagy, P. (2006). Evidence-Based Psychodynamic Psychotherapies.
Psychodynamic Diagnostic Manual. Silve Spring, MD, Alliance of Psychoanalytic
Organisations: 765-818.
Fonagy, P., G. Gergely, et al. (2002). Affect Regulation, Mentalization and the Development of the Self. NY, Other Press. Fonagy, P., A. Roth, et al. (2005). "Psychodynamic psychotherapies: evidence-
based practice and clinical wisdom." Bulletin of the Menninger Clinic 69(1): 1-
58.
Gabbard, G. O., S. G. Lazar, et al. (1997). "The economic impact of
psychotherapy: a review." American Journal of Psychiatry 154: 147-55.
Leichsenring, F. (2005). "Are psychodynamic and psychoanalytic therapies effective? A review of empirical data." International Journal of Psychoanalysis 86: 841-68.
Lyons-Ruth, K. (1995). "Broadening our conceptual frameworks: can we
reintroduce relational strategies and implicit representational systems to the
study of psychotherapy?" Developmental Psychology 31(3): 432-436.
Main, M. and E. Hesse (2001). "Attachment narratives and attachment accross
the lifespan." Journal of the Americal Psychoanalytic Association 48.
Read, J. e. a. (2005). "Childhood trauma, psychosis and schizophrenia: a litrature review with theoretical and clinical implications." Acta Psychiatrica Scandanavica 112: 330-350.
Schore, A. (1999). Affect Regulation and the Origin of the Self : The Neurobiology of Emotional Development, Lawrence Erlbaum. Siegel, D. (1999). The Developing Mind - How relationships and the brian interact to shape who we are. NY, London, Guilford. Stern, D. (1985). The Interpersonal World of the Infant - A View from Psychoanalysis and Developmental Psychology, Basic Books. Surgenor, D. L. (2006). Consultation on the form of regulation of psychotherapy under the Health Practitioners Competency Assurance Act 2003: letter to Ryan McLean, Ministry of Health, NZ Government, Psychologists' Board.

Source: http://www.nztaa.org.nz/Is-Psychotherapy-Any-Good.pdf