Purpose: We compared the Neurological Disorders Depression Inventory for Epilepsy (NDDI-E) with the Patient Health Questionnaire-9 (PHQ-9) according to different antiepileptic drug (AED) loads in people with epilepsy (PWE).
Methods: Depression and suicidality were assessed with the Mini International Neuropsychiatric Interview (MINI), the NDDI-E, and the PHQ-9. A receiver operating characteristic (ROC) curve analysis was used. Sensitivity and specificity of the NDDI-E and PHQ-9 were compared between an AED load <2 and ≥2.
Results: Of 213 participants included, 22.5 % were diagnosed with current depression by the MINI, and 9 % had a moderate to severe risk of suicide. Using the cutoff of 9.5 and 13.5 for the PHQ-9 and NDDI-E, respectively, in the total group, the PHQ had a higher sensitivity (91.7 %) but lower specificity (75.8 %) than the NDDI-E (66.7 % and 87.3 %, respectively) (p < 0.001). In a group with an AED load <2, the sensitivity and specificity did not differ between the PHQ-9 and NDDI-E. In a group with an AED load ≥2, however, the PHQ-9 had a higher sensitivity (90.9 %, p < 0.063) but lower specificity (65.2 %, p < 0.001) than the NDDI-E (68.2 % and 81.2 %, respectively). For screening of suicide risk, the sensitivity and specificity were not different between the PHQ-9 and NDDI-E in all tested groups. The negative predictive values of both instruments were above 95 % for suicide risk.
Conclusion: In patients with a low AED load, the sensitivity and specificity did not differ between the PHQ-9 and NDDI-E. In contrast, in patients with a high AED load, the PHQ-9 had a poor specificity whereas the NDDI-E had a poor sensitivity.
Keywords: Depression; Epilepsy; NDDI-E; PHQ-9; Questionnaire; Suicidality; Validation.
Copyright © 2019 British Epilepsy Association. Published by Elsevier Ltd. All rights reserved.