Abstracts

Rationale: The aim of this study is to assess and compare the efficacy of the Beck Depression Inventory (BDI), Hospital Anxiety and Depression Scale Depression Subscale (HADS-D), and the Korean version of the Neurological Disorders Depression Inventory for Epilepsy (K-NDDI-E) for identifying suicidality in people with epilepsy (PWE).Methods: A Total of 97 PWE attending our epilepsy outpatient completed screening tests for depression including the BDI, HADS-D, and K-NDDI-E. According to the known cutoff scores for depression, 50 possibly depressed patients in at least one test underwent further structured diagnostic interview, the Mini International Neuropsychiatric Interview (MINI-Plus) for assessment of depressive disorder and suicide risk. We compared the clinical and demographic characteristics between low and moderate-to-severe suicide risk group according the score in suicide section of MINI-plus.Results: In terms of MINI-plus diagnoses, 22 patients had moderate-to-high risk of suicide and 12 patients had major depressive episode, dysthymia or other depressive disorder. PWE with moderate-to-severe suicide risk had younger onset age, higher BDI and K-NDDI-E scores than low suicide risk group. For identifying moderate-to-severe risk of suicide, HADS-D was inappropriate with small area under the ROC curves compared with BDI and K-NDDI-E. A cutoff point of 11 on the K-NDDI-E, the same cutoff for screening depression, showed excellent sensitivity of 100.0% and reasonable specificity of 68.7% with high negative predictive value of 100%. BDI with a cutoff point of 14 had strong sensitivity of 90.9% and acceptable specificity of 71.4% with good negative predictive value of 90.9%. The same cutoff for depression on BDI indicated low sensitivity of 68.2% and acceptable specificity of 78.6%. A comparison of the AUC showed statistical difference between BDI and K-NDDI-E, which AUCs were significantly larger than that of HADS-D.Conclusions: K-NDDI-E is more recommended than HADS-D for screening both depression and suicide risk with consistent cutoff value of 11.