Abstracts

Optimal Cut-Off Score for Diagnosing Major Depressive Disorder With the Neurological Depression Disorder Inventory for Epilepsy (NDDI-E): A Meta-Anlaysis

Abstract number : 3.373
Submission category : 11. Behavior/Neuropsychology/Language / 11A. Adult
Year : 2018
Submission ID : 501047
Source : www.aesnet.org
Presentation date : 12/3/2018 1:55:12 PM
Published date : Nov 5, 2018, 18:00 PM

Authors :
Do-Hyung Kim, Gyeongsang National University Hospital; Young-Soo Kim, Gyeongsang National University Hospital; Tae-Won Yang, Gyeongsang National University Hospital; Jongsoo Kang, Gyeongsang National University Hospital; and Oh-Young Kwon, Gyeongsang Nati

Rationale: The Neurological Depression Disorder Inventory for Epilepsy (NDDI-E) is quick and useful inventory for screening depression in people with epilepsy (PWE). This inventory has been validated in various language versions, and is commonly used in epilepsy clinics. The cutoff score for screening major depressive disorder (MDD) has been reported with the range of 11-16. However, it is difficult which cutoff score should be selected when the NDDI-E is used with the languages without validation. The aim of this study is to find optimal cutoff point of NDDI-E score for detecting MDD by combining the data from previous studies.  Methods: We searched MEDLINE, EMBASE, Cochrane Library, Web of Science, and SCOPUS to identify studies. Original researches which tested the accuracy of NDDI-E for detecting MDD in adult PWE were recruited for this meta-analysis. We included the studies when MDD was diagnosed by the gold standard structural interviews such as Mini International Neuropsychiatric Interview. In addition, we included the studies for this meta-analysis only when the studies provided sufficient information to obtain contingency tables. Two-by-two cross-tables were made with MDD and NDDI-E using data provided in the primary studies finally included. ROC analyses were performed with these data. We obtained summary receiver operating characteristic curve at each cutoff score from 9 to 20. We also obtained odd ratio between PWE with MDD and those without MDD by meta-analysis using a fixed effects model. Results: We identified 13 validation studies conducted in outpatient epilepsy clinics settings. The area under the curve (AUC) value was more than 0.830 at three cutoff scores from 13 to 15. Seven, 4, and 10 studies provided data for the each cutoff score 13, 14 and 15 respectively. The total number of PWE was 1078, 712, and 1386, respectively at each cutoff score after combining data from the primary studies. Cutoff score 14 and 15 were most valid for detecting  MDD. The AUC value was 0.838 (SE = 0.0173, p <0.01) and 0.839 (SE = 0.0713, p < 0.01) at cutoff score 14 and 15 respectively. When screening MDD with the cutoff score 14, the sensitivity was 79.3% (95% confidence interval [CI]: 72.3 - 85.2) and specificity was 88.3% (95% CI: 85.3 - 90.9). When screening MDD with the cutoff score 15, the sensitivity was 79.2% (95% CI: 74.1 - 83.8) and specificity  was 88.5% (95% CI: 86.5 - 90.3). Conclusions: The NDDI-E has an acceptable diagnostic properties for detecting MDD for cut-off scores between 13 and 15. Of these, the cutoff score 14 and 15 may be most optimal cutoff scores. This information will be a helpful reference when screening MDD using NDDI-E in the areas where NDDI-E has not been validated for their languages. Funding: We do not have any funding for this study.