To the Editor.—
While in agreement with the findings that physicians have a significantly higher suicide rate than the general population (228:323, 1974), we seriously question the methodology. The use of obituaries in The Journal has been discredited as a reliable research method,1-3 and it has been found to underreport physician suicides by as much as 50%.3Drs. Steppacher and Mausner's claim that women physicians have a suicide rate three times that of a comparable female population needs to be confirmed by other investigators utilizing more reliable and comprehensive sampling methods, such as reviewing all the death certificates of physicians filed in a state for a given time period. Whatever difficulties are posed by any suicide research, particularly on special groups such as physicians, sampling bias is by far the most critical in limiting the generality of results.