Posts Tagged ‘DSM’
It has been postulated (for example see Persons & Miakami) that psycho diagnostic evaluation is an important phase in the treatment of psychopathology. However there is a fierce battle going on between those that argue for a stepped-care model in mental health care and those that argue that we should rather look at the specific needs of an individual (matched care – could not find a suitable link -why?).
Connected to that is the question of how much information about the client you need in order to start treatment and if there is an amount of client information that is enough (that does not enhance the quality of the treatment plan any more).
If you can suffice with the DSM or ICD classification for treatment planning – why should you waste time on making an elaborate functional analysis of the client’s problems? However research (Persons & Miakami) has shown that treatment is not successful in about 40% of the cases and can be improved if diagnosticians make a functional analysis. This can sometimes help to focus treatment on the most important factor that keeps the client from getting healthy.
On the other hand Garb (1998 – it’s a book – for an overview read his 2004 paper: Clinical Judgment and Decision Making) has reported that lively details can influence clinicians to the degree that they make a wrong judgment about treatment decisions.
Thus we would like to know how much information should a therapist have in order to make a good treatment planning? Below I give you are plan for study 1 (total of 5 studies) with which we would like to start-up this line of research. It’s still in the planning phase and things have been changed around a lot, but we have found the direction I guess. I would like to ask you some feedback on this research.
Do you think this study answers the research question?
What do you think is the minimum amount of information that still feels “natural”?
What kind of disorders and empirically supported treatments (EST) would you chose?
What do we need to take care of when writing the vignettes?
Thanks a lot for your feedback. This is an experiment in itself: Can the global community of researchers and therapist help on improving this research idea? Can we work together – even though we will never meet? Or will someone steal this idea and conduct the research themselves? I am very thankful for your critical comments!
Research Proposal Study 1
Study 1 addresses the question whether increasing amounts of diagnostic information, from classification only to classification plus extensive case formulation, changes clinicians’ treatment plans. On the basis of the literature and in line with national and international treatment guidelines, three frequently occurring DSM-IV Axis-I (clinical syndromes) disorders are selected for which familiar first-choice psychological ESTs are available. Using real patient files as a starting point, clinical vignettes are constructed, each one presenting a patient suffering from one of the selected DSM-IV disorders. These three vignettes are rewritten in such a way as to generate three Diagnostic Classification (DC) versions: DC, DC+, and DC++. Starting with the most complete DC++ versions, increasingly information is deleted to come to the DC+ and the DC versions. DC vignettes comprise 5-8 lines. From the DC++ versions information is removed, save: 1. demographic information (e.g., sex, age, marital status, children, current job), 2. treatment history (in all vignettes: ‘moderately successful pharmacological treatment only’), 3. current complete DSM-IV diagnosis, and 4. a recommended EST. DC+ vignettes are twice as long as DC vignettes and comprise 10-16 lines. They contain all the information that is available in DC vignettes, plus further anamnestic and psychodiagnostic information (e.g., family background, life history, personality). On the other hand, all lively details in the DC++ versions have been removed, in line with findings of Garb (1998) that clinicians are biased towards lively details in patients’ life histories. DC++ vignettes comprise 20-32 lines. They contain lively details originating from the patient files of real patients that are used as a starting point. The vignettes are piloted and tested for their ecological validity (cf. De Kwaadsteniet, Krol, & Witteman, 2008; Hutschemaekers, Tiemens, & Kaasenbrood, 2005; Witteman & Koele, 1999).
By contacting mental health institutes, 90 psychologists or psychiatrists are sought who have been involved in the intake and diagnosis of outpatients at least once a week for at least five years, and who are willing to participate in a study on diagnostic decision-making carried out by email. Each participant receives 3 vignettes. The participants are told that these three diagnostic reports were made by experienced clinicians of a large mental health institute. The participants are asked per vignette to address the following three issues: 1. What do you think of the recommended psychological treatment?, 2. Do you have recommendations for additional treatment or interventions?, and 3. Do you want to add contra-indications for certain interventions? Using a Latin Square Design for the selection of vignettes, it is ensured that for each participant all three patients and each of the three versions (DC, DC+ or DC++) are represented, and that all combinations (patients x versions) occur equally as often. For each of the three patient vignettes 30 DC, 30 DC+, and 30 DC++ versions are available. Power tables show that with alphas set at .05 and n = 30 in each cell, moderate to large effects can be identified with a power of .80 (Garssen & Hornsveld, 1992). When comparing the vignette versions without further regard to the patients, there are 90 versions in each cell, enough power to detect small effects.
We expect that the more diagnostic information is left out of the vignettes, the lower the variability in treatment suggestions of the participants. The hypotheses are that the evaluations of DC++ versions compared to evaluations of DC+ and DC versions, and evaluations of DC+ versions compared to DC versions show: 1. more deviations from recommended, first choice EST, and 2. larger numbers of indicated and contra-indicated treatment suggestions.
Please give us some feedback – we are also thankful for questions and remarks about what is clear/not clear and about what you like/dislike! Also you might want to suggest a flash title for our research!
According to Lars Bergman there are several reasons that keep computerized systems from producing the correct DSM-classification. The most important point might be that diagnosis in psychiatry is just to complex for a fully structured system to reach a good level of validity. Dynamic systems and appropriate statistics might be necessary. According to Dreyfus the intuition of the expert is an important part of the diagnostic process. This would also mean that the (paper version of the) SCID does not yield the expected validity. Other reasons are related to methodological problems of the study at hand: Most physicians used the computer version of the SCID for the first time. They also received only limited training and thus the computerized process might have interfered with their normal skills. Also the rule-following and logical step-by-step procedures designed for non-experts might have influenced validity.
In medicine, diseases are classified according to the physiological problem that is presumed to have caused the disease. Classification is guided by advanced biological and chemical theories that have proven to have explanatory value. For mental illnesses that is a different story, because the working mechanism is mostly unknown and there is no unified scientific framework that can explain the abnormal processes. Mental health professionals for different schools might explain any given symptom according to a different theory. In that light the DSM was created to be purely atheoretical and give a minimal framework on which everyone involved could agree upon. The resulting manual has helped communication about mental illnesses and has fostered comparable research in the field of mental health. However it is questionable if the structure of the DSM is compatible with the reality of humans, who (for the most) are driven by the unstoppable urge to make sense of their environment. Indeed it was found in many studies that therapists often classify without adhering to the specific rules laid out by the DSM. They are guided by internal rules, that work in a different way than the DSM checklist. For example therapists regard some symptoms as more important for a specific diagnose than others (concept of centrality). The DSM on the other hand usually assigns equal weights to all factors. Because the concept of centrality has profound influence on memory and other cognitive processes the therapist’s personal theory plays an important role in the classification. I believe that the DSM should be regarded as a tool supporting communication that allows discussion about the different theories at hand. It can and will not replace the theories people have. If a theory based classification system is available (for a specific disease) it should be used next to or instead of the DSM.