The Hypomania Checklist (HC) is a 20-item questionnaire, which is easy to fill in and designed to help clinicians collect data for diagnosing bipolar disorder. This tool could be very useful in primary care where type II bipolar disorder is under-diagnosed, however, to date no suitable cut-off score correlating with a high probability of bipolar II disorder diagnosis has yet been validated. Method In a French clinico-epidemiological multi-center survey (EPIDEP) a national sample of patients with DSM-IV major depressive episode (MDE) was recruited and assessed at admittance and four weeks later. Diagnoses of unipolar or bipolar disorder were made according to a semi-structured interview adapted from the DSM-IV. In addition, the HC and questionnaires on affective temperament were administered at the second interview. In the analyses, the diagnostic accuracy was computed in terms of sensitivity, specificity, predictive positive value and predictive negative value, by varying cut-off scores on the HC. The Receiver Operating Characteristic (ROC) statistical technique was used to compare the diagnostic value of HC with the semi-structured interview adapted from the DSM-IV. Results Of the 493 patients with a MDE DSM-IV diagnosis, 468 filled in the HC, from which the six following groups were formed: strict unipolar disorder (UP, N = 201), bipolar I disorder (BP-I, n=39), bipolar II disorder (BP-II, N = 141), patients with mania or hypomania secondary to an antidepressant treatment ( N = 51), cyclothymia ( N = 14) and hyperthymia ( N = 22). Comparing the BP-II patient group ( N = 141) with the strict UP group ( N = 201) the most discriminating HC score was 9, which identified 81% of patients correctly, with a sensitivity of 86.5, a specificity of 77.1, a predictive positive value of 72.6 and a predictive negative value of 89.1. Some cases identified as cyclothymic and hyperthymic temperaments by the affective temperament questionnaire but meeting DSM-IV criteria for major depressive disorder were included in the unipolar group ( N = 237). The same score of 9 was validated, identifying a percentage of patients with correct diagnosis of unipolar depression of 78.3%, a sensitivity of 86.5, a specificity of 73.4, a predictive positive value of 66.0 and a predictive negative value of 90.1. If patients with mania or hypomania secondary to an antidepressant treatment were included as a subgroup of BP-II, a score of ten appeared as the most relevant, with a percentage of patients with correct diagnosis of 79.0 %, a sensitivity of 80.2, a specificity of 78.1, a predictive positive value of 74.8 and a predictive negative value of 83.0. ROC curves confirmed these values. Lastly when BP-I patients ( N = 39) were compared to the strict UP group ( N = 201) the most discriminating HC score was 11, with a percentage of patients with correct diagnosis of 86.3%, a sensitivity of 74.4, a specificity of 88.6, a predictive positive value of 55.8 and a predictive negative value of 94.7, but the BP-I group was too small to validate the score of 11. Conclusions These results indicate that a score of 9 on the HC is highly correlated with a BP-II diagnosis (and a score of 10 if patients with mania induced by antidepressants are considered as BP-II), and suggest that a wider use of the HC in primary care associated with strong GP/Psychiatrist networks could improve the detection, and with appropriate treatment, the prognosis of Bipolar II disorder.
Read full abstract