A STUDY OF PSYCHIATRIC DIAGNOSIS

Abstract

Ten diagnosticians, sitting as a group, individually diagnosed a total of 91 patients. The patient group were first admissions, aged 15-59, to a public mental hospital. Utilizing 12 standard diagnostic categories, the overall likelihood of a second opinion agreeing with the first was 57%. There was considerable variation in reliability by diagnostic categories. Groupings of categories gave better reliability. Neither the overuse or underuse of diagnostic categories by individual diagnosticians, nor diagnostician "expectancy," were significant factors in diagnostic disagreement in this study. Selection by "high" certainty of diagnosis gave a modest improvement in diagnostic reliability. The opinion is expressed that future studies must focus, not just on diagnosis, but the diagnostic process.