Issues in research into therapy * 50 years on from Eysenck’s 1952 paper, there is little resemblance between therapy then & now * Old research issues still present * New issues: * Evidence based practice * Role of meta analysis in building theory * Research into clinical utility or cost-effectiveness rather than efficacy Outcome research since 1952 * Outcome initially judged on pre / post treatment comparison and therapist ratings * Eysenck identifies spontaneous remission problem * So comparison / control groups * no treatment, waiting list, placebo to control for spontaneous remission, placebo & Hawthorne effects * Manualised interventions * Multiple / blind / objective ratings But tightly controlled outcome studies bring their own problems * Internal validity relies on rigorous control of variables * short term analog study? * External (ecological) validity prioritises meaning over rigour * messy & ethically compromised clinical settings over longer term? Changing research questions * Does therapy work? (efficacy) * Which therapies work best? (comparative efficacy) * Paul’s (1967) call for a matrix of therapies / conditions / other variables – a precursor of evidence-based therapy * (Cost) effectiveness / clinical utility * If…...then…...pragmatic specificity – plugging gaps in the missing matrix * eg. early intervention in psychosis Changing funding & political climate * Demand for social goods is rising faster than for consumer goods * Governments supply social goods, clashes with political pressure for lower taxes * = market pressure to raise productivity * = managed care, dose-effect curve * = research into clinical utility or cost- effectiveness rather than efficacy Dose effect curve Hansen, Lambert & Foreman 2002, Clinical Psychology: Science and Practice Review * Eysenck (’52) study based on data not now acceptable * 1960 study reached same conclusion with better data * Radical improvements in efficacy research methodology * Allegedly little influence of research on clinical practice * But eclecticism, integration, manualisation show influence * Evidence-based practice a way of getting research evidence to influence clinical practice * Meta analysis began to reverse Eysenck’s conclusions and show a consistent effect of therapy Evidence based practice * Development of evidence based practice guidelines in USA and UK may lead to prescriptive managed care * Is this a good thing? * Nature of evidence accepted * Positivist epistemology and drug trial gold standard may act as a straight-jacket? * Tendency to favour cognitive behavioural & outcome focussed approaches & work against process orientations? * Parry (2000) rejects special pleading Marzillier (2004) The myth of evidence-based practice * Generalised outcomes oversimplify people’s lives & the intricacies of therapy * Bentall (2003) undermines Kraepelin-based diagnosis, therapists ought not to buy into the medical model – should always start with the individual, each case is unique, vast differences within the same category * Research generally lacks ecological validity * Therapy at heart is a personal transaction / relationship * Argues that clients need to tell their stories, the importance of non-specific factors Efficacy research * Systematic evaluation in a controlled research environment (Barlow ’96). * Typical study: Control condition (wait list, placebo), random assignment, control of variables, single-blinded, restrictive inclusion criteria, manualised with specifically trained staff, outcomes are short-term, targeted at symptom change. * Aim at measurable effects of specific interventions &, above all, replicability Effectiveness research * Applicability / feasibility of intervention with established efficacy in local clinical setting (Barlow ’96) * Use people in need of treatment regardless of specific diagnosis, co-morbidity, length of illness. * Treatment method, frequency, duration & assessment follows clinical need rather than research design * Outcome measures broader - quality of life, degree of disability * Aim at external validity & generalisability Meta analysis and building theory * Schmidt (’92) argues that the use of effect size and meta analysis renders the traditional experimental trial partly redundant * Meta analysis enables systematic & authoritative reviews of efficacy & allows a test of hypotheses about the conditions under which a stronger, weaker or non-existent effect will be observed. * Foundation for evidence based practice guidelines? Comparison & control issues revisited * Control conditions * Are they still necessary? Ethical problem, bias problem – participants may not engage enough, volunteers may be very different to real patients * Several types – is a placebo really possible? * Negative and positive effects * Alternative is to compare to therapy of known effectiveness Issues in methodology * Is the traditional, experimental, randomised control trial (RCT) still the gold standard? * Manualisation of treatment - helps internal validity & standardisation but limits natural development of therapy * Well defined groups * Screening to get a group of patients who are homogenous in diagnosis & severity may lead to response bias. Alternative methodologies? * Consumer satisfaction (Seligman) * Clinical audit * See Curtis-Jenkins (2002) Good money after bad? In: Feltham, What’s the good of counselling etc. * Action-research * Qualitative / constructionist approaches to understanding the meaning of therapy as an outcome mediator at a cultural level and the construction of meaning in therapy at an individual level * See McLeod (2003) on the social meaning of counselling, pps 33-36 Summary: Problems & directions in outcome research * See review by Nathan, Stewart & Dolan (2000) * On meta analysis see: * Schmidt, F.L. (1992) What do data really mean? Research findings, meta analysis, and cumulative knowledge in psychology. American Psychologist. 47, 10, 1173-1181 * Key issues: * To research efficacy or effectiveness? * To integrate findings and find consensus? * To understand more about process?