Abstract
Background: Nurses play a crucial role in the routine assessment of depression. However, brief depression assessment tools-necessary for busy clinical settings-have not been evaluated to identify depression in women with histories of intimate partner violence.
Objective: To evaluate the utility of two 5-question subsets from the Center for Epidemiologic Studies-Depression (CES-D) scale in detecting depressive symptoms in women with abuse histories.
Methods: The sample comprised 448 women involved in police- or court-reported incidents of intimate partner violence who completed a questionnaire used to assess depression with the 20-item CES-D scale. Receiver operating characteristic (ROC) analysis was used to identify score thresholds for two 5-question subsets from the CES-D for detecting (a) depressive symptoms and (b) severe depressive symptoms. Depression prevalence was estimated using score thresholds identified in the ROC analysis. The discriminating ability of the CES-D question subsets was also evaluated.
Results: Using thresholds identified in the ROC analyses, sensitivities ranged from .94 to .95 according to the CES-D question subsets for depressive symptoms and .97 to .98 for severe depressive symptoms. Specificity ranged from .73 to .87. Depression prevalence according to the 20-item CES-D was 84% for depressive symptoms and 67% for severe depression. Depression prevalences were 81%-84% (depressive symptoms) and 72% (severe depressive symptoms) using the CES-D question subsets. The two CES-D question subsets were comparable in their ability to identify minor and severe depressive symptoms, using the 20-item score as the gold standard (area under the curve range = .96-.97).
Discussion: Two brief question subsets were effective in identifying depression and can be used by nurses to assess depression in women with histories of abuse.