To the Editor.—
Dr Jonathan Stewart of the New York State Psychiatric Institute has brought to our attention an error that appeared in our article.1 In Table 7, there was a misprint; the P value produced when unipolar depression is defined as need for hospitalization should have been.0931 rather than.0031. In addition, in that table, a colon was inadvertently inserted for a period in the typesetting process. The numbers in the last column of that table should have read 1.2, 1.4, etc, rather than 1:2, 1:4, etc. The insertion of colons erroneously gives the impression of higher unipolar-bipolar ratios than actually exist.We appreciate having an opportunity to clarify these points.