Premium
Validating screening tools for depression in epilepsy
Author(s) -
Fiest Kirsten M.,
Patten Scott B.,
Wiebe Samuel,
Bulloch Andrew G.M.,
Maxwell Colleen J.,
Jetté Nathalie
Publication year - 2014
Publication title -
epilepsia
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 2.687
H-Index - 191
eISSN - 1528-1167
pISSN - 0013-9580
DOI - 10.1111/epi.12754
Subject(s) - gold standard (test) , epilepsy , depression (economics) , hospital anxiety and depression scale , patient health questionnaire , comorbidity , population , medicine , anxiety , psychiatry , physical therapy , depressive symptoms , environmental health , economics , macroeconomics
Summary Objective Depression is a common comorbidity of epilepsy, and its timely identification in persons with epilepsy is essential. The use of screening tools to detect depression is common in epilepsy, but some scales in current use have not been validated using a gold standard in this population. The present study aims to validate three commonly used depression‐screening scales and assess new cut points for scoring in those with epilepsy. Methods Persons with epilepsy (n = 300) from the only epilepsy clinic in a large urban health region completed questionnaires (e.g., sociodemographics, adverse event profile) and three depression‐screening tools (Hospital Anxiety and Depression Scale [ HADS ]; Patient Health Questionnaire [ PHQ ]‐9 and PHQ ‐2). One hundred eighty‐five patients participated in a gold‐standard structured clinical interview to assess depression. The diagnostic accuracy of the depression scales was assessed comparing a variety of scoring cut points to the gold‐standard diagnosis of depression. Results The prevalence of current depression in this population, according to the gold‐standard, was 14.6%. The scale with the highest sensitivity (84.6%) was the HADS with a cut point of 6 and the scale with the highest specificity (96.2%) was the PHQ ‐9 algorithm scoring method. Overall, the PHQ ‐9 at a cut point of 9 and the HADS at a cut point of 7 resulted in the greatest balance of sensitivity and specificity (area under the curve: 88% and 90%, respectively). Significance The PHQ ‐9 at a cut point of 9 and the HADS at a cut point of 7 had the best overall balance of sensitivity and specificity. However, for screening purposes the PHQ ‐9 algorithm method is ideal (optimizing specificity), whereas for case finding the HADS at a cut point of 6 performed best (optimizing sensitivity). Appropriate scale cut points should be chosen based on the study's goals and available resources. Disease‐specific scale cut points are recommended for future studies assessing depression in persons with epilepsy.