Medical decision making in antipsychotic drug choice for schizophrenia
Hamann J, Langer B, Leucht S, Busch R and Kissling W;
Commented by , 28 Jul 2000
Aim of the study
How psychiatrists make treatment decisions is poorly understood. In other medical fields the interdisciplinary research field medical decision making that attempts to explore physicians’ decision making has attracted a lot of interest in recent years. In the field of psychiatry this research seems to be very important, too.
For example, although an increasing number of treatment guidelines recommend new generation antipsychotics (NGA) as first line drugs for those with schizophrenia, prescription data suggest that at least in Germany first generation antipsychotics (FGA) are still used for many patients. The aim of the current study therefore was to explore psychiatrists’ decision making process when choosing among first generation and new generation antipsychotic drugs.
Method
100 physicians (50 in private practice and 50 from 8 hospitals) in the region of southern Bavaria were questioned with semi-structured interviews on antipsychotic drug choice (FGA or NGA) in patients with schizophrenia or schizoaffective disorder according to ICD-10. Hospital physicians were asked concerning their next two patients to be discharged from hospital and psychiatrists in private practice were questioned with regard to the last two patients they had either initiated or switched treatment with an antipsychotic.
Data on the characteristics of patients (age, gender, severity of illness etc.) and of physicians were documented. After identifying variables that may influence the treatment decision with t-tests and chi-square tests, the two patients of each physician were randomly assigned to a "training sample" and to a "confirmative sample". Both samples were then analysed using multiple logistic regression.
Results
Of the variables identified to be of potential interest by t-tests and chi-square tests, physicians' age, poor compliance and duration of the illness statistically significantly influenced the decision in favour of the prescription of a first generation antipsychotic according to multiple logistic regression. However, when the same variables were entered in the confirmative sample, only the physician’s age showed a statistically significant influence on drug choice.
Discussion
The main result of this study was that physicians’ age was the strongest and the only robust predictor for the choice between FGAs and NGAs. First of all it is remarkable that a physician related factor (physicians' age) and not patient related factors influenced drug choice most, although treatment should be oriented towards the patient and not towards the doctor. Furthermore, the study suggests that physicians prefer those drugs they are more familiar with (older, more experienced physicians prefer FGAs, younger, less experienced physicians prefer SGAs).
On the one hand it may be reasonable to prescribe the drug that one can handle best. On the other hand such an experience-based prescribing of older doctors may lead to an undertreatment with new and better tolerable compounds. A strength of the study was that the doctors were questioned on "real" patients, while other frequently used approaches such as case vignettes and mail surveys may be biased by social desirability and are decisions in hypothetical situations. In conclusion, the study underlines the need for medical decision making research in psychiatry.