Tardive dyskinesia (TD), an abnormal movement disorder seen in patients taking neuroleptic drugs, afflicts large numbers of institutionalized and chronically ill psychiatric patients for whom these drugs are the prime treatment medium. The long-term objective and aim of the proposed project is to characterize the natural history of TD by defining what happens to TD over time--does it persist, change or resolve and what are the patient characteristics and treatment history variables that relate to the outcomes? Parkinsonism-like symptoms (PK) and akathisia (AKA) will be included in the analysis. The proposed work will address the public health dimensions of TD by providing a clear picture of the size and makeup of the TD pool over time. Psychiatric inpatients (N=301) evaluated (1979-1983) by a single rater for TD, PK and AKA form the study population. The proposed project will reevaluate these patients using the same rater, instruments and procedures. The interval length between Evaluation 1 (E1) and Evaluation 2 (E2) will range from 3-8 years. Current status of the 301 patients is as follow: inpatients--211; outpatients--42; discharged--30; deceased--18. It is estimated that 89% (268) can be reached for reevaluation. In addition to side effect evaluations, a diagnostic interview will be conducted and a BPRS administered. Psychotropic and antiparkinson inpatient medication use for the E1/E2 interval will be extacted from a computerized drug ordering system. Outpatient medication use will be taken from clinical records. Dependent variables for TD will be formed from the Simpson Abbreviated Dyskinesia Scale and the Abnormal Involuntary Movement Scale. Independent variables will characterize: course of illness, psychiatric status, diagnosis, sex, age, PK, AKA, coexisting disorders and all facets of medication use for the E1/E2 interval. T-tests, chi-squares, and simple linear regressions will be used for data screening. The large sample size will allow for significant variables to be tested in logistic multiple and multiple linear regression models, stratifying and covarying for sex and age, to sort out relative contributions and interactions. Data primarily will be analyzed in two levels. Level I analyses will answer TD outcome questions, including only those persons who reached a positive TD designation at E1 with the TD status and severity at E2 as the dependent measures. Level II analyses will include those persons evaluated at E2 with the TD status at E2 as the dependent measure. These Level II data will, in comparison to the E1 data, define what happens to the size and makeup of the TD pool over time.
