Skeptics doubt that the high rates of depression found in community epidemiologic surveys represent clinical disorders. They question whether the conditions reported by respondents to these household surveys may in fact reflect trivial or transient problems.

Two new studies reported in the March Archives of General Psychiatry demonstrate that far from overstating the prevalence of depression, the surveys may underestimate rates. Both studies compare survey interview schedules, which rely on self-report, with similar schedules administered by clinicians.

As part of a follow-up of the 1981 Epidemiologic Catchment Area (ECA) study in Baltimore, psychiatrists examined 349 individuals who had responded to questions from lay interviewers using the Diagnostic Interview Schedule (DIS). The psychiatrists followed the Schedules for Clinical Assessment in Neuropsychiatry (SCAN) in making their diagnoses.

William W. Eaton and his colleagues at Johns Hopkins University and a component of the World Health Organization in Kobe, Japan, found that agreement in the results from the two types of interview was "only fair." Many respondents "repeatedly failed to report DIS symptoms attributed to life crises or medical conditions," reported Eaton and his colleagues.

Similar results were found by Jane M. Murphy, Ph.D., and her associates at Harvard University and Dalhousie University in Nova Scotia. As part of their continuing research in "Stirling County" on the Atlantic coast of Canada, they compared diagnostic interview schedules administered by clinicians and nonprofessional interviewers.

Most cases diagnosed by two lay-administered methods were corroborated by clinicians, who found many additional cases. Murphy and her colleagues cautioned that because the validity of the various instruments has not yet been adequately demonstrated, these results should be considered tentative.