A psychometric evaluation of the 16-item PHQ-ADS concomitant anxiety and depression scale in the UK biobank using item response theory.

Autor: Pflanz CP; Dementias Platform UK, Department of Psychiatry, Warneford Hospital, University of Oxford, Oxford OX3 7JX, United Kingdom. Electronic address: patrick.pflanz@psych.ox.ac.uk., Gallacher J; Dementias Platform UK, Department of Psychiatry, Warneford Hospital, University of Oxford, Oxford OX3 7JX, United Kingdom., Bauermeister S; Dementias Platform UK, Department of Psychiatry, Warneford Hospital, University of Oxford, Oxford OX3 7JX, United Kingdom.
Jazyk: angličtina
Zdroj: Journal of affective disorders [J Affect Disord] 2024 Feb 15; Vol. 347, pp. 335-344. Date of Electronic Publication: 2023 Nov 22.
DOI: 10.1016/j.jad.2023.11.067
Abstrakt: Background: The Patient Health Questionnaire Anxiety and Depression Scale (PHQ-ADS) provides a reliable and valid measure of concomitant depression and anxiety. However, research on its psychometric efficiency and optimal scale length using item-response theory (IRT) has not been reported. This study aimed to optimize the length of the PHQ-ADS scale without losing information by discarding items that were a poor fit to the IRT model.
Methods: The UK Biobank is a large cohort study designed to investigate risk factors for a broad range of disease. PHQ-ADS data were available from n = 152,826 participants (age = 55.87 years; SD = 7.73; 56.4 % female), 30.4 % of the entire UK Biobank sample. Psychometric properties of the PHQ-ADS were investigated using a 2-parameter IRT and Mokken analysis. Item statistics included discrimination, difficulty and Loevinger H coefficients of monotonicity.
Results: In the entire 16-item scale, item discrimination ranged from 1.40 to 4.22, with the item 'worrying' showing the highest level of discrimination and the item 'sleep disturbance' showing the lowest. Mokken analysis showed that the 16-item PHQ-ADS scale could be reduced to a 7-item scale without loss of test information. The reduced scale comprised mainly items measuring cognitive-affective symptoms of anxiety/depression, whereas items measuring somatic symptoms were discarded. The revised scale showed high discrimination and scalability.
Limitations: Findings are limited by the use of cross-sectional data that only included the baseline online questionnaire, but not other waves.
Conclusions: IRT is a useful technique for scale reductions which serve the clinical and epidemiological need to optimize screening questionnaires to reduce redundancy and maximize information. A reduced-item 7-item PHQ-ADS scale reduces the response burden on participants in epidemiological research settings, without loss of information.
Competing Interests: Declaration of competing interest SB, CPP and JG declare no competing interests.
(Copyright © 2023 The Author(s). Published by Elsevier B.V. All rights reserved.)
Databáze: MEDLINE
Externí odkaz: