Differential Item Functioning of the HADS and PHQ-9: An investigation of age, gender and educational background in a clinical UK primary care sample

Research output: Contribution to journalArticle

19 Citations (Scopus)

Abstract

Background
The Patient Health Questionnaire (PHQ-9) and Hospital Anxiety and Depression Scale (HADS) are commonly used measures in clinical practice and research. It is important that such scales measure the trait they purport to measure and that the impact of other measurement artefacts is minimal. Differential item functioning of these scales by gender, educational background and age is currently assessed.

Methods
Severity of depression and anxiety symptoms were measured in primary care patients referred to mental health workers using the PHQ-9 and HADS. Each scale was assessed for Differential Item Functioning (DIF) and Differential Test Function (DTF) by gender, educational background and age. Minimum n per analysis=895. DIF was assessed with Mantel's ¿2, Liu-Agresti cumulative common odds ratio (LA LOR) and the standardised LA LOR (LA LOR-Z). DTF was assessed in relation to ¿2.

Results
PHQ-9, HADS Depression Sub-scale (HADS-D) and HADS Anxiety Subscale (HADS-A) lacked bias in terms of gender and educational background (¿2<0.07). However, both PHQ-9 and HADS-D exhibited bias with regard to age: PHQ-9 ¿2=0.103 (medium effect); HADS-D ¿2=0.214 (large effect). PHQ-9 items exhibiting DIF by age covered: anhedonia, energy and low mood. HADS-D items exhibiting DIF by age covered psychomotor retardation and interest in appearance.

Limitations
No assessment of other potential DIF contributors was made.

Conclusions
PHQ-9, HADS-D and HADS-A generally do not exhibit bias for gender and educational background. However bias was observed in PHQ-9 and HADS-D for age. Caution should be exercised interpreting scores both in clinical practice and research.

Original languageEnglish
Pages (from-to)262-268
Number of pages7
JournalJournal of Affective Disorders
Volume147
Issue number1-3
Early online date4 Dec 2012
DOIs
Publication statusPublished - May 2013

Fingerprint

Primary Health Care
Anxiety
Depression
Odds Ratio
Anhedonia
Sexism
Research
Artifacts
Mental Health
Health

Keywords

  • Depression
  • Anxiety
  • Questionnaires
  • bias

Cite this

@article{49d2c41fa4af41a98e8dd3d621f232fb,
title = "Differential Item Functioning of the HADS and PHQ-9: An investigation of age, gender and educational background in a clinical UK primary care sample",
abstract = "Background The Patient Health Questionnaire (PHQ-9) and Hospital Anxiety and Depression Scale (HADS) are commonly used measures in clinical practice and research. It is important that such scales measure the trait they purport to measure and that the impact of other measurement artefacts is minimal. Differential item functioning of these scales by gender, educational background and age is currently assessed. Methods Severity of depression and anxiety symptoms were measured in primary care patients referred to mental health workers using the PHQ-9 and HADS. Each scale was assessed for Differential Item Functioning (DIF) and Differential Test Function (DTF) by gender, educational background and age. Minimum n per analysis=895. DIF was assessed with Mantel's ¿2, Liu-Agresti cumulative common odds ratio (LA LOR) and the standardised LA LOR (LA LOR-Z). DTF was assessed in relation to ¿2. Results PHQ-9, HADS Depression Sub-scale (HADS-D) and HADS Anxiety Subscale (HADS-A) lacked bias in terms of gender and educational background (¿2<0.07). However, both PHQ-9 and HADS-D exhibited bias with regard to age: PHQ-9 ¿2=0.103 (medium effect); HADS-D ¿2=0.214 (large effect). PHQ-9 items exhibiting DIF by age covered: anhedonia, energy and low mood. HADS-D items exhibiting DIF by age covered psychomotor retardation and interest in appearance. Limitations No assessment of other potential DIF contributors was made. Conclusions PHQ-9, HADS-D and HADS-A generally do not exhibit bias for gender and educational background. However bias was observed in PHQ-9 and HADS-D for age. Caution should be exercised interpreting scores both in clinical practice and research.",
keywords = "Depression, Anxiety, Questionnaires, bias",
author = "Cameron, {Isobel M.} and Crawford, {John R.} and Kenneth Lawton and Reid, {Ian C.}",
year = "2013",
month = "5",
doi = "10.1016/j.jad.2012.11.015",
language = "English",
volume = "147",
pages = "262--268",
journal = "Journal of Affective Disorders",
issn = "0165-0327",
publisher = "Elsevier",
number = "1-3",

}

TY - JOUR

T1 - Differential Item Functioning of the HADS and PHQ-9

T2 - An investigation of age, gender and educational background in a clinical UK primary care sample

AU - Cameron, Isobel M.

AU - Crawford, John R.

AU - Lawton, Kenneth

AU - Reid, Ian C.

PY - 2013/5

Y1 - 2013/5

N2 - Background The Patient Health Questionnaire (PHQ-9) and Hospital Anxiety and Depression Scale (HADS) are commonly used measures in clinical practice and research. It is important that such scales measure the trait they purport to measure and that the impact of other measurement artefacts is minimal. Differential item functioning of these scales by gender, educational background and age is currently assessed. Methods Severity of depression and anxiety symptoms were measured in primary care patients referred to mental health workers using the PHQ-9 and HADS. Each scale was assessed for Differential Item Functioning (DIF) and Differential Test Function (DTF) by gender, educational background and age. Minimum n per analysis=895. DIF was assessed with Mantel's ¿2, Liu-Agresti cumulative common odds ratio (LA LOR) and the standardised LA LOR (LA LOR-Z). DTF was assessed in relation to ¿2. Results PHQ-9, HADS Depression Sub-scale (HADS-D) and HADS Anxiety Subscale (HADS-A) lacked bias in terms of gender and educational background (¿2<0.07). However, both PHQ-9 and HADS-D exhibited bias with regard to age: PHQ-9 ¿2=0.103 (medium effect); HADS-D ¿2=0.214 (large effect). PHQ-9 items exhibiting DIF by age covered: anhedonia, energy and low mood. HADS-D items exhibiting DIF by age covered psychomotor retardation and interest in appearance. Limitations No assessment of other potential DIF contributors was made. Conclusions PHQ-9, HADS-D and HADS-A generally do not exhibit bias for gender and educational background. However bias was observed in PHQ-9 and HADS-D for age. Caution should be exercised interpreting scores both in clinical practice and research.

AB - Background The Patient Health Questionnaire (PHQ-9) and Hospital Anxiety and Depression Scale (HADS) are commonly used measures in clinical practice and research. It is important that such scales measure the trait they purport to measure and that the impact of other measurement artefacts is minimal. Differential item functioning of these scales by gender, educational background and age is currently assessed. Methods Severity of depression and anxiety symptoms were measured in primary care patients referred to mental health workers using the PHQ-9 and HADS. Each scale was assessed for Differential Item Functioning (DIF) and Differential Test Function (DTF) by gender, educational background and age. Minimum n per analysis=895. DIF was assessed with Mantel's ¿2, Liu-Agresti cumulative common odds ratio (LA LOR) and the standardised LA LOR (LA LOR-Z). DTF was assessed in relation to ¿2. Results PHQ-9, HADS Depression Sub-scale (HADS-D) and HADS Anxiety Subscale (HADS-A) lacked bias in terms of gender and educational background (¿2<0.07). However, both PHQ-9 and HADS-D exhibited bias with regard to age: PHQ-9 ¿2=0.103 (medium effect); HADS-D ¿2=0.214 (large effect). PHQ-9 items exhibiting DIF by age covered: anhedonia, energy and low mood. HADS-D items exhibiting DIF by age covered psychomotor retardation and interest in appearance. Limitations No assessment of other potential DIF contributors was made. Conclusions PHQ-9, HADS-D and HADS-A generally do not exhibit bias for gender and educational background. However bias was observed in PHQ-9 and HADS-D for age. Caution should be exercised interpreting scores both in clinical practice and research.

KW - Depression

KW - Anxiety

KW - Questionnaires

KW - bias

U2 - 10.1016/j.jad.2012.11.015

DO - 10.1016/j.jad.2012.11.015

M3 - Article

VL - 147

SP - 262

EP - 268

JO - Journal of Affective Disorders

JF - Journal of Affective Disorders

SN - 0165-0327

IS - 1-3

ER -