Abstract
Objective: To validate the factor structure of the Patient Health Questionnaire Anxiety and Depression Scale (PHQ-ADS)- which is a composite measure of depression and anxiety using the Patient Health Questionnaire-9 and Generalised Anxiety Disorder Scale (GAD-7), in a sample of haemodialysis patients.
Method: Screening data (n = 182) used to select entry into a feasibility study of an online cognitive-behavioural therapy intervention for distress in dialysis patients were analysed here. Structural validity of the PHQ-ADS was evaluated using confirmatory factor analysis (CFA), assessing alternative models including a bi-factor model. In the bi-factor model all items from the PHQ-9 and GAD-7 (16-items in total) were loaded onto a general distress factor. Respective items of the PHQ-9 and GAD-7 were specified as subgroup factors. Omega-hierarchical was calculated to indicate the level of saturation of a multidimensional scale by a general factor. Construct validity was determined against the Brief Illness Perception Questionnaire.
Results: A bi-factor PHQ-ADS model had good fit to the data (chi-square = 96.1, p = 0.26, CFI = 0.99; TLI = 0.99; RMSEA = 0.02). The general distress factor accounted for approximately 84% of the explained variance (omega-h = 0.90). Distress scores were significantly higher in females compare with males. There was a significant association between distress and negative illness perceptions (r = 0.58, p < 0.01).
Conclusions: The PHQ-ADS appears to have good structural validity in haemodialysis patients and is sufficiently unidimensional to warrant the use of a total distress score. A full psychometric analysis of the PHQ-ADS in a larger sample of dialysis patients is warranted.
Method: Screening data (n = 182) used to select entry into a feasibility study of an online cognitive-behavioural therapy intervention for distress in dialysis patients were analysed here. Structural validity of the PHQ-ADS was evaluated using confirmatory factor analysis (CFA), assessing alternative models including a bi-factor model. In the bi-factor model all items from the PHQ-9 and GAD-7 (16-items in total) were loaded onto a general distress factor. Respective items of the PHQ-9 and GAD-7 were specified as subgroup factors. Omega-hierarchical was calculated to indicate the level of saturation of a multidimensional scale by a general factor. Construct validity was determined against the Brief Illness Perception Questionnaire.
Results: A bi-factor PHQ-ADS model had good fit to the data (chi-square = 96.1, p = 0.26, CFI = 0.99; TLI = 0.99; RMSEA = 0.02). The general distress factor accounted for approximately 84% of the explained variance (omega-h = 0.90). Distress scores were significantly higher in females compare with males. There was a significant association between distress and negative illness perceptions (r = 0.58, p < 0.01).
Conclusions: The PHQ-ADS appears to have good structural validity in haemodialysis patients and is sufficiently unidimensional to warrant the use of a total distress score. A full psychometric analysis of the PHQ-ADS in a larger sample of dialysis patients is warranted.
| Original language | English |
|---|---|
| Pages (from-to) | 15-19 |
| Journal | GENERAL HOSPITAL PSYCHIATRY |
| Volume | 50 |
| Early online date | 28 Sept 2017 |
| DOIs | |
| Publication status | Published - 2018 |