What Gets Surveyed in a Psychiatric Survey?

This research tests and analyzes the operation of one of the earliest psychiatric screening instruments. It remains a most suitable instrument for surveying a general population with lay interviewers. The instrument used was an abbreviated version of the MacMillan Index first used in the Stirling County studies. The research design was to compare the scores obtained from a general population to the scores obtained by a group of clinic patients drawn from the same population. Results show that the index clearly discriminates between patients and the undiagnosed parent population. 1) Thirteen of the 15 questions discriminate between the two groups far beyond chance (p <.0001). 2) A principal factor analysis indicates the operation of two independent factors in the responses of non patients, one characterized by impaired work and social functioning and the other by symptoms of general anxiety. 3) Discriminant function analysis reveals four key questions, two dealing with impaired function, one with somatic distress, and one with fear of a “nervous breakdown.” The authors conclude that the economy, ease, and speed of administration of the MacMillan Index, combined with its demonstrated ability to discriminate between known patients and a general population, make it an excellent instrument for psychiatric epidemiological studies, although its lack of specificity impairs its use as a diagnostic instrument. They urge its further development and analysis.