Screening for depression in arthritis populations: an assessment of differential item functioning in three self-reported
- PDF / 453,177 Bytes
- 11 Pages / 595.276 x 790.866 pts Page_size
- 73 Downloads / 174 Views
Screening for depression in arthritis populations: an assessment of differential item functioning in three self-reported questionnaires Jinxiang Hu1
•
Michael M. Ward1
Accepted: 23 May 2017 Ó Springer International Publishing Switzerland 2017
Abstract Purpose To determine if persons with arthritis differ systematically from persons without arthritis in how they respond to questions on three depression questionnaires, which include somatic items such as fatigue and sleep disturbance. Methods We extracted data on the Centers for Epidemiological Studies Depression (CES-D) scale, the Patient Health Questionnaire-9 (PHQ-9), and the Kessler-6 (K-6) scale from three large population-based national surveys. We assessed items on these questionnaires for differential item functioning (DIF) between persons with and without self-reported physician-diagnosed arthritis using multiple indicator multiple cause models, which controlled for the underlying level of depression and important confounders. We also examined if DIF by arthritis status was similar between women and men. Results Although five items of the CES-D, one item of the PHQ-9, and five items of the K-6 scale had evidence of DIF based on statistical comparisons, the magnitude of each difference was less than the threshold of a small effect. The statistical differences were a function of the very large sample sizes in the surveys. Effect sizes for DIF were similar between women and men except for two items on the Patient Health Questionnaire-9. For each questionnaire,
Electronic supplementary material The online version of this article (doi:10.1007/s11136-017-1601-x) contains supplementary material, which is available to authorized users. & Jinxiang Hu [email protected]; [email protected] 1
Intramural Research Program, National Institute of Arthritis and Musculoskeletal and Skin Diseases, National Institutes of Health, Bethesda, MD, USA
DIF accounted for 8% or less of the arthritis-depression association, and excluding items with DIF did not reduce the difference in depression scores between those with and without arthritis. Conclusions Persons with arthritis respond to items on the CES-D, PHQ-9, and K-6 depression scales similarly to persons without arthritis, despite the inclusion of somatic items in these scales. Keywords Differential item functioning MIMIC Depression Arthritis
Introduction Self-reported questionnaires are widely used to assess depressive symptoms or serious psychological distress in clinical studies and epidemiological surveys. Several reliable and valid questionnaires, such as the 20-item Center for Epidemiologic Studies Depression (CES-D) scale, have been developed for this purpose [1, 2]. Newer shorter scales have been designed that assess more specifically symptoms and signs of depression represented in the Diagnostic and Statistical Manual of Mental Disorders criteria or the International Classification of Disease criteria. These measures include the Patient Health Questionnaire-9 (PHQ-9) and the Kessler-6 (K-6) scale [3–5]. All th
Data Loading...
