Factors related to prescription dosage in Japanese psychiatric hospitals

A point prevalence study of prescription dosage was made among psychiatric inpatients in Japan with the following aims: (i) to analyze whether total chlorpromazine equivalent dose was related to patient characteristics or hospital and physician characteristics; and (ii) to analyze whether the usage of the two major categories of antipsychotics, butyrophenone and phenothiazine, were related to their expected pharmacological reactions. A comprehensive database of prescription contents and characteristics of patients, physicians and hospitals was developed from a survey of 1674 patients in 18 psychiatric hospitals. Using the tree-based model analysis, factors related to the total chlorpromazine equivalent dose were analyzed. Factors included hospital ownership, staffing levels, physician specialty and patient length of stay, symptoms etc. The same analysis was then done for the two major categories of antipsychotics, in order to evaluate whether their expected pharmacological reactions were related to their usage. Factors related to the total chlorpromazine equivalent dose were age of patient, presence of hallucination, age of first psychiatric hospitalization, and change(s) in attending physician (while hospitalized). For the chlorpromazine equivalent dose of the butyrophenone category, the clinical factors related were abnormality in thought and hallucination, and for the phenothiazine category, it was bizarre behavior. Total chlorpromazine equivalent doses had no relationship with hospital or physician characteristics, and appeared to be determined by patient age and psychiatric symptoms. Doses in the butyrophenone category were related to pathological experiences and that in the phenothiazine category to behavioral aspects.