Background: The Beck Depression Inventory-II (BDI-II) is often used to assess depressive symptoms among stroke patients, but more evidence is needed regarding its psychometric properties in this population. The purpose of this study was to assess the BDI-II׳s psychometric properties using a Rasch model application in a sample of patients 6 months after a first clinical stroke.
Methods: Data were collected prospectively from patient medical records and from questionnaires (with assistance if needed) as a part of a longitudinal study of poststroke fatigue. Data from the 6-month follow-up were used in this analysis. The sample consisted of 106 patients with first-ever stroke recruited from two Norwegian hospitals between 2007 and 2008. Depressive symptoms were measured with the BDI-II. Rasch analysis was used to assess the BDI-II׳s psychometric properties in this sample.
Results: Five BDI-II items did not demonstrate acceptable goodness-of-fit to the Rasch model: items 10 (crying), 16 (changes in sleep), 17 (irritability), 18 (changes in appetite), and 21 (loss of interest in sex). If these 5 items were removed, the resulting 16-item version not only had fewer items, it also had better internal scale validity, person-response validity, and person-separation reliability than the original 21-item version in this sample of stroke survivors.
Limitations: The study did not include a clinical evaluation of depression.
Conclusion: A 16-item version of the BDI-II, omitting items 10, 16, 17, 18 and 21, may be more appropriate than the original 21-item BDI-II for use as a unidimensional measure of depression in patients following first-ever stroke.
Keywords: BDI-II; Depression; Psychometrics; Rasch analysis; Stroke; Validity.
Copyright © 2014 The Authors. Published by Elsevier B.V. All rights reserved.