Optimal cutoff score of the Neurological Disorders Depression Inventory for Epilepsy (NDDI-E) for detecting major depressive disorder: A meta-analysis

Abstract

Background and purpose: The Neurological Disorders Depression Inventory for Epilepsy (NDDI-E) is a useful inventory for screening major depressive disorder (MDD) in people with epilepsy (PWE). The cutoff score for detecting MDD has been reported with the range of >11 to >16. The aim of this study was to find optimal cutoff score of the NDDI-E for MDD detection by combining the raw data from previous studies. Methods: We searched MEDLINE, EMBASE, Cochrane Library, Web of Science, and SCOPUS to identify proper studies. Original researches, which tested the accuracy of NDDI-E for MDD detection in adult PWE, were recruited. We included the studies in which MDD was diagnosed by a gold standard structural interview, the Mini International Neuropsychiatric Interview (MINI). In addition, we included only the studies providing enough information for meta-analysis: number of PWE with MDD, number of total PWE, and sensitivity (Se) and specificity (Spe) for each cutoff score. After collecting data from included studies, we performed a diagnostic test accuracy (DTA) meta-analysis using bivariate model. Results: We identified 13 validation studies conducted in outpatient epilepsy clinic setting. As summary estimates of test accuracy measures, the Se, Spe, and diagnostic odds ratio (DOR) of NDDI-E for MDD detection were 0.81, 0.84, and 22.48, respectively. The analysis using the multiple thresholds model showed that the NDDI-E score of 13.2 was the best fit for MDD detection. When analyzing only with the seven data sets of the cutoff score >13, the Se, Spe, and DOR were 0.87, 0.80, and 25.72, respectively. Conclusions: The optimal NDDI-E cutoff score for MDD detection is >13. The information provided by this DTA meta-analysis will be a useful reference for applying NDDI-E in geographic areas where no NDDI-E validation studies have been conducted for their languages. (C) 2018 Elsevier Inc. All rights reserved.