Salma A Ayis1, Luis Ayerbe2, Mark Ashworth3, Charles DA Wolfe4. 1. School of Population Health and Environmental Sciences, Faculty of Life Sciences and Medicine King's College, London, UK; National Institute for Health Research (NIHR) Biomedical Research Centre, Guy's and St Thomas' NHS Foundation Trust, London, UK; National Institute for Health Research (NIHR) Collaboration for Leadership in Applied Health Research and Care (CLAHRC) South London at King's College Hospital NHS Foundation Trust, UK. Electronic address: salma.ayis@kcl.ac.uk. 2. Blizard Institute. Centre for Primary Care and Public Health, Barts and The London School of Medicine and Dentistry, Queen Mary University of London, UK. 3. School of Population Health and Environmental Sciences, Faculty of Life Sciences and Medicine King's College, London, UK. 4. School of Population Health and Environmental Sciences, Faculty of Life Sciences and Medicine King's College, London, UK; National Institute for Health Research (NIHR) Biomedical Research Centre, Guy's and St Thomas' NHS Foundation Trust, London, UK; National Institute for Health Research (NIHR) Collaboration for Leadership in Applied Health Research and Care (CLAHRC) South London at King's College Hospital NHS Foundation Trust, UK.
Abstract
BACKGROUND: Variations have been reported in the number of underlying constructs and choice of thresholds that determine caseness of anxiety and /or depression using the Hospital Anxiety and Depression scale (HADS). This study examined the properties of each item of HADS as perceived by stroke patients, and assessed the information these items convey about anxiety and depression between 3 months to 5 years after stroke. METHODS: The study included 1443 stroke patients from the South London Stroke Register (SLSR). The dimensionality of HADS was examined using factor analysis methods, and items' properties up to 5 years after stroke were tested using Item Response Theory (IRT) methods, including graded response models (GRMs). RESULTS: The presence of two dimensions of HADS (anxiety and depression) for stroke patients was confirmed. Items that accurately inferred about the severity of anxiety and depression, and offered good discrimination of caseness were identified as "I can laugh and see the funny side of things" (Q4) and "I get sudden feelings of panic" (Q13), discrimination 2.44 (se = 0.26), and 3.34 (se = 0.35), respectively. Items that shared properties, hence replicate inference were: "I get a sort of frightened feeling as if something awful is about to happen" (Q3), "I get a sort of frightened feeling like butterflies in my stomach" (Q6), and "Worrying thoughts go through my mind" (Q9). Item properties were maintained over time. LIMITATIONS: Approximately 20% of patients were lost to follow up. CONCLUSION: A more concise selection of items based on their properties, would provide a precise approach for screening patients and for an optimal allocation of patients into clinical trials.
BACKGROUND: Variations have been reported in the number of underlying constructs and choice of thresholds that determine caseness of anxiety and /or depression using the Hospital Anxiety and Depression scale (HADS). This study examined the properties of each item of HADS as perceived by strokepatients, and assessed the information these items convey about anxiety and depression between 3 months to 5 years after stroke. METHODS: The study included 1443 strokepatients from the South London Stroke Register (SLSR). The dimensionality of HADS was examined using factor analysis methods, and items' properties up to 5 years after stroke were tested using Item Response Theory (IRT) methods, including graded response models (GRMs). RESULTS: The presence of two dimensions of HADS (anxiety and depression) for strokepatients was confirmed. Items that accurately inferred about the severity of anxiety and depression, and offered good discrimination of caseness were identified as "I can laugh and see the funny side of things" (Q4) and "I get sudden feelings of panic" (Q13), discrimination 2.44 (se = 0.26), and 3.34 (se = 0.35), respectively. Items that shared properties, hence replicate inference were: "I get a sort of frightened feeling as if something awful is about to happen" (Q3), "I get a sort of frightened feeling like butterflies in my stomach" (Q6), and "Worrying thoughts go through my mind" (Q9). Item properties were maintained over time. LIMITATIONS: Approximately 20% of patients were lost to follow up. CONCLUSION: A more concise selection of items based on their properties, would provide a precise approach for screening patients and for an optimal allocation of patients into clinical trials.
Authors: Anne Vinggaard Christensen; Jane K Dixon; Knud Juel; Ola Ekholm; Trine Bernholdt Rasmussen; Britt Borregaard; Rikke Elmose Mols; Lars Thrysøe; Charlotte Brun Thorup; Selina Kikkenborg Berg Journal: Health Qual Life Outcomes Date: 2020-01-07 Impact factor: 3.186