Madness Explained

by Richard P. Bental

  The recent development of a new group of atypical neuroleptics adds a final twist to this unfortunate story. The first of these drugs, clozapine, was synthesized in the 1960s37 but was quickly discontinued when it was discovered that about one in 200 patients receiving it suffered from agranulocitosis (as a consequence, eight elderly patients in Finland died during an early clinical trial). However, interest in the drug was revived when it was realized that its low profile of extra pyramidal side effects made it an attractive treatment for patients suffering from tardive dyskinesia. Procedures were therefore developed to allow patients to be closely monitored so that they could be quickly withdrawn from the drug if signs of agranulocitosis became evident. A clinical trial conducted by John Kane, Herbert Meltzer and their colleagues in the United States, published in 1988, seemed to confirm that clozapine could be taken safely, caused fewer extra pyramidal side effects than chlorpromazine and was an effective treatment for patients who failed to respond to conventional drugs.38

  Pharmaceutical companies quickly realized that drugs such as clozapine might provide an opportunity to increase dramatically the profits they made from selling their products to psychiatric services. (As most of the conventional neuroleptics are out of patient, they can be manufactured as ‘generic drugs’ by any company, keeping profit margins very tight.)In the last few years, therefore, several more atypical neuroleptics have been licensed and marketed as being kinder and more effective than conventional anti-psychotic medication. Given that these drugs are extremely expensive to tax payers (in the UK) and insurance companies (in the USA) it is obviously important to establish whether these claims are justified. A good starting point is John Kane’s 1988 trial.

  In order to select patients who had failed to respond to typical neuroleptics, Kane and his colleagues required that anyone entering their study had experienced at least three attempts at conventional treatment with doses equivalent to one gram of chlorpromazine a day or higher. Of the 268 patients who eventually took part, half were randomly assigned to treatment with clozapine. The remaining patients were assigned to treatment with doses of chlorpromazine up to a (literally) staggering 1.8 grams per day. The patients were then followed up for six weeks, during which time their psychotic symptoms and side effects were monitored. The researchers claimed that those treated with clozapine experienced less severe psychotic symptoms, and also fewer extra pyramidal side effects, than those treated with chlorpromazine (although statistically significant differences were only observed on one of the two side-effect measures used in the study).

  Considerable effort was required in order to carry out the experiment. Herbert Meltzer has estimated that, because of the difficulty in finding suitable patients and in monitoring them adequately, it cost the pharmaceutical company Sandoz over $5 million to complete.39 However, despite this investment, the results of the study are almost impossible to interpret because the comparison drug, chlorpromazine, was given at a dose that was much too high to be optimally therapeutic. It has been shown that patients treated with unnecessarily high doses of conventional neuroleptics sometimes improve when their drug doses are reduced.40 It is therefore possible that the patients in the control group would have done as well as those patients given clozapine had they been given a more sensible amount of chlorpromazine. Furthermore, given the dose they actually received, it is astonishing that much larger differences in side effects were not observed.

  A recent analysis of data from clinical trials of the atypicals, published in the British Medical Journal, has confirmed that the effectiveness of the new drugs has been systematically exaggerated by studies that have compared them to inappropriate doses of conventional medication. British psychiatrist John Geddes and his colleagues analysed data from fifty-two studies involving a total of 12,649 patients, in which the experimental medication was either clozapine or one of four other recently introduced atypicals (risperidone, olanzapine, quetiapine, or sertindole).41 When the trials were divided into those in which control patients had received a daily dose equivalent to 12 milligrams or less of haloperidol (equivalent to about 300 milligrams of chlorpromazine), and those in which the comparison dose was greater, no differences in outcome were observed in the former studies, but the atypicals appeared more effective and more acceptable to patients in the latter. The only evidence of superiority for the atypicals that remained when the dose of the comparison drug was optimal was in extrapyramidal effects, which remained less frequent and less severe in those receiving the new medications. However, even this benefit seems less than impressive when it is realized that the atypical medications can cause other types of side effects in abundance (olanzapine, for example, is notorious for causing weight gain).

  Addressing Patients’ Psychological Needs

  I expect that some psychiatrists will react to the account I have given of neuroleptic treatment by assuming that I am motivated by some kind of ideological hostility to medical treatment. Certainly, there are some mental health professionals – usually psychologists, but sometimes nurses or even psychiatrists – who believe that the use of psychiatric drugs is always wrong, and that neuroleptics are mere ‘chemical straitjackets’.42 This is not my position. What I am arguing for is a balanced appraisal of the benefits and costs of neuroleptic treatment, based on scientific evidence. There is no doubt that neuroleptics are a useful therapeutic tool if prescribed to the right patients in sensible doses, but there is equally no doubt that they can be harmful if used in excess. The failure to recognize the very real limitations of this kind of treatment is just as short sighted as the blanket rejection of drug treatment of any kind.

  The way forward in the drug treatment of psychosis is therefore to find ways of targeting medication more accurately to meet the needs of patients. As it is not possible to predict neuroleptic response in advance of treatment, the only way that this can be achieved is by adopting a systematic policy of ‘suck it and see’. To do this effectively, prescribers will have to work closely with patients in order to monitor therapeutic benefits and side effects, and will also need the courage to withdraw patients completely from their drugs if they are obviously failing to benefit. Without proper research into drug-free treatment (which is almost impossible in medically dominated psychiatric services) it is difficult to know how many patients are best treated without drugs of any kind, but the number is likely to be many more than are treated this way at present. (In the only study so far conducted to address this question, carried out by Loren Mosher in the USA in the 1970s, it was found that most first-episode patients treated without drugs but given very intensive psychological support did just as well as drug-treated patients at two-year follow-up.43 However, this important study has never been properly replicated and, given the dominance of Kraepelinian thinking in psychiatry, is not likely to be repeated in the foreseeable future.)

  Just as the Kraepelinian paradigm has encouraged clinicians to rely on neuroleptic drugs, it has discouraged the provision of psychological treatments. Further discouragement has been provided by the results of clinical trials of intensive psychoanalytic treatment (based on the ideas of Freud and his followers). This type of psychotherapy encourages patients to explore difficult emotional issues with therapists who offer interpretations of their thoughts and feelings, but little in the way of concrete advice. Although Freud himself was sceptical about the value of this approach in the treatment of psychosis, it was adapted for this use by a number of American psychoanalysts, most notably Harry Stack Sullivan,44 a charismatic psychiatrist who was influential in the 1930s and 1940s, and whose theories of mental illness were a blend of ideas garnered from Freud and social psychology.45

  Unfortunately, when clinical trials have been conducted to assess the effects of psychoanalytic treatment on positive symptoms the results have been almost universally negative.46 The most rigorous study of this kind was carried out in the early 1980s by Alfred Stanton, John Gunderson and their colleagues in Boston, USA.47 One hundred and sixty-five patients we
re offered therapy by highly experienced clinicians twice a week for up to two years. By six months, all but sixty-nine had dropped out, itself an indication that the therapy failed to meet the needs of the majority of patients. After two years, a control group of patients who had received simple supportive counselling (emotional support and advice about practical difficulties) were no more symptomatic, and had spent less time in hospital and more time in employment, than those psychoanalytically treated patients who had persisted to the end. Naturalistic long-term follow-up studies of patients who had received intensive psychoanalytic therapy in specialist hospitals, published at about the same time, yielded equally discouraging results.48

  Although psychoanalytic theory has at times led to useful insights into the behaviour and experience of patients, analysts have been reluctant to integrate their ideas with equally important insights gained from scientific research. As a consequence, the theory has not led to a coherent understanding of the origins of psychosis, or experimentally testable theories about the psychological processes responsible for complaints such as delusions and hallucinations. Perhaps we therefore should not be surprised that treatment based on psychoanalytic principles offers few benefits for psychotic patients. By contrast, the psychological models discussed throughout this book have been based on studies that have followed the normal conventions of scientific investigation. They should therefore be robust enough to guide the development of novel interventions. In fact, two types of treatment that are consistent with these models seem to offer tangible benefits for patients and their families.

  In Chapter 16 I described evidence that family relationships can affect the long-term course of psychotic difficulties. Patients returning from hospital to live with relatives who are either critical and hostile or emotionally over-controlling are more likely to relapse than patients living in less stressful circumstances. It did not take long for the researchers who made this discovery to realize that providing assistance to patients’ families might therefore benefit both the families and patients. In a paper published in 1982, Julian Leff and his colleagues reported the effects of a nine-month treatment programme, in which the relatives of schizophrenia patients were educated about psychosis, took part in support groups, and met with therapists who tried to help them improve their ability to resolve family conflicts.49 Twenty-four patients were followed up, initially for nine months and then, in a subsequent report, two years after treatment had begun.50 Patients whose families had received the intervention had a much lower rate of relapse than patients in the control group (8 per cent versus 50 per cent at nine months and 20 per cent versus 78 per cent at two years).

  This finding has since been repeated many times, for example by Ian Falloon and his colleagues in southern California,51 by Gerry Hogarty and others in Pittsburgh,52 and by my colleagues Nick Tarrier and Christine Barrowclough in Salford, near Manchester.53 A recent systematic review identified twenty-five clinical trials published in English- and German-language journals in which relatives had been included in treatments for schizophrenia patients.54 In some of these studies, benefits were maintained many years after treatment had ended.55 Overall, the patients whose relatives participated in these programmes benefited from a 20 per cent reduction in relapses during the first year after treatment commenced, a benefit similar in magnitude to that observed in trials comparing neuroleptic medication with placebos. Much less effort has been made to investigate the value of this kind of treatment for bipolar patients and their families, but at least one study has reported encouraging results.56

  Whereas family interventions attempt to ameliorate the kind of stress that makes psychotic symptoms worse, more recently developed cognitive behaviour therapy (CBT) interventions are designed to influence the cognitive processes that give rise to symptoms. These interventions have been modified from earlier techniques that were developed for the treatment of depression by Aaron Beck and others in the USA.57 Unlike psychoanalytic therapy, CBT is targeted at precisely specified difficulties that are identified by the patient and the therapist together. (Often, the first thing that they do together is draw up a ‘problem list’.) It involves the patient and therapist collaborating in a very practical way to find solutions to the patient’s problems. (At the end of each session, patients usually take away homework assignments so that they can test out new ideas or try out novel solutions to their problems that emerge during the course of discussion with the therapist.) The core idea of CBT is that patients can learn to reflect on their own thoughts and beliefs, find ways of testing their validity, and, if necessary, substitute more helpful thoughts and beliefs. Patients receiving CBT for depression, for example, learn to recognize when their attributions for events are unnecessarily pessimistic, to replace them with more realistic appraisals of events, and to find proactive ways of solving their problems.

  Many of the psychologists who first began to experiment with CBT for psychosis were also involved in some of the basic research that I have described earlier in this book. For some reason (probably because they were relatively uncommitted to the neoKraepelinian dogma) nearly all were British. Paul Chadwick, whose ideas about paranoia were briefly considered in Chapter 13, reported some early case studies of deluded patients who had been successfully encouraged to reconsider their abnormal beliefs.58 Gill Haddock, whose work we briefly touched on in Chapter 14, investigated different methods of helping hallucinating patients to identify the source of their inner speech.59 Max Birchwood, working with Paul Chadwick, studied the effects of encouraging hallucinating patients to question their beliefs about their voices.60 However, psychiatrists David Kingdon (now at the University of Southampton) and Douglas Turkington (now at the University of Newcastle) and clinical psychologist Nick Tarrier (at the University of Manchester) probably deserve most credit for integrating these ideas into comprehensive therapeutic strategies.

  Kingdon and Turkington proposed a general framework for working with psychotic patients, which they described as a normalizing strategy.61 The idea behind this approach is to demystify psychotic experiences and make them seem less frightening, for example by pointing out the similarities between hallucinations or paranoia and more mundane mental states, or by explaining to patients that these experiences are much more common than is often realized. Nick Tarrier’s approach was initially named coping skills enhancement, because it aimed to help patients learn better ways of coping with their experiences, but was soon expanded to incorporate various strategies for addressing patients’ delusional ideas or beliefs about their voices.62

  Nick Tarrier published the first randomized controlled clinical trial of cognitive behaviour therapy for schizophrenia patients. At six-month follow-up, patients whose delusions and hallucinations had not responded to neuroleptic medication, and who had been offered CBT, had shown a greater improvement in their complaints compared to patients who were taught simpler problem-solving skills.63 The trial was soon followed by another small-scale study, carried out by Philippa Garety and her colleagues at the Institute of Psychiatry in London, who found greater improvements in delusional beliefs in chronically ill patients treated with CBT than in patients who did not receive this kind of therapy.64 In the last few years, larger trials have begun to appear in scientific and medical journals, nearly all of which have supported these initial results. At the time of writing, seven trials have appeared, including further studies by David Kingdon, Douglas Turkington and their colleagues,65 by investigators at the Institute of Psychiatry,66 and by Nick Tarrier and his colleagues.67 Several others have been completed and are awaiting publication. One of these is the SoCRATES(Study of Cognitive-Realignment Therapy in Early Schizophrenia) trial conducted by Shôn Lewis, Nick Tarrier, Gill Haddock, myself and others, in which we have found that a very brief (five-week) intervention offered to patients when they become acutely ill for the very first time results in modest benefits at eighteen-month follow-up.68 Another is a study conducted by Andrew Gumley and his colleagues in Scotland, who have found
that a very brief intervention with patients who have been ill for some time can dramatically reduce the likelihood that they will relapse.69 Although bipolar patients have, again, been largely overlooked in these developments, three small-scale studies have provided preliminary evidence that CBT may be an effective method of reducing the risk of manic episodes.70

  None of these findings suggests that CBT or family therapy is a panacea for madness. Despite our best efforts, some patients fail to respond to psychological treatment and it is often forgotten that less glamorous interventions, for example providing opportunities for marginalized and isolated patients to find work, or solving their accommodation problems, can have effects on quality of life that are just as dramatic as any that can be achieved by psychologists or psychiatrists.

  Nonetheless, very real progress has been made in the psychological treatment of madness over the last decade or so. Furthermore, it is unlikely that the best that can be accomplished today is the last word on what is possible. Just as pharmaceutical companies do not assume that today’s medications cannot be bettered, we should not assume that existing results represent some kind of limit on what psychological interventions can do for people afflicted with madness.