COMPUTER DIAGNOSIS--EPIDEMIOLOGIC APPLICATIONS

本研究涉及五台微机的开发与评价

This research involves the development and evaluation of five microcomputer administered versions of the NIMH-Diagnostic Interview Schedule (DIS). The DIS is being used and studied in many epidemiologic, research and clinical settings where it has been shown to have good reliability. Time to administer the interview is acceptable for most research uses but addition of new diagnoses will make the full interview too long for some applications. Scoring of the interview already requires a computer. Specific Aim #1 is to provide all interviews on a microcomputer system capable of running them in an acceptable manner and at a low cost. The IBM PC-XT has been selected and will be capable of storing more than 400 full interviews as well as generating summaries minutes after data collection is completed. The DIS interviews will be programmed in MUMPS standard and a second version will also run under C/PM-86 or DOS 2.C. Specific Aim #2 is the preparation and evaluation of five DIS interviews for microcomputer administration. Three computer interviews will be prepared to prompt human interviewers: Full DIS; Shortened DIS; and Quick Entry DIS. Two direct patient-computer interviews will also be prepared: Full DIS and Shortened DIS. All five interviews will use the same scoring program and can generate summaries which will identify the data source (interviewer or patient) and nature of the interview (full, short or quick entry). Emphasis will be placed on the shortened versions. Evaluations will be conducted in Madison, Wisconsin and St. Louis, Missouri with more than 300 patients and will permit comparisons between the different methods for four major diagnostic categories (schizophrenia, affective disorders, anxiety disorders and drug abuse or antisocial personality disorders). The final product of this research will be a set of computer programs that can be made widely available at low cost. These programs will reduce time and cost of training interviewers, largely eliminate the need for interviewers in the patient-computer version and permit updates to be made uniformly from floppy discs.