Rasch Analysis of the Dimensional Structure of the Hospital Anxiety

1
Rasch Analysis of the Dimensional Structure of
the Hospital Anxiety Depression Scale (HADS)

• Adam Smith
• Cancer Research-UK Psychosocial Clinical
  Practice Group, St. Jamess University Hospital,
  Leeds
• Institute of Psychological Sciences, University
  of Leeds

2
Groups Aims

• Patient-centred assessment in clinical practice
  for patients with cancer
• Evaluation of quality of life
• Social problems
• Psychosocial morbidity

3
Overview

• Collaboration
• CR UK groups in Edinburgh (Prof. M. Sharpe) and
  Leeds (Prof. P. Selby)
• Rasch analysis gt 5 yrs
• Mainly alongside CTT, but more recently item-bank
  development
• Rasch Analysis of the HADS
• Development of a screening instrument

4
Psychological distress cancer

• Psychological distress in cancer patients
• Prevalence
• Anxiety disorders 7 - 23 (Stark House, 2000)
• Depression 7 - 47 (Sellick Crooks, 1999)

5
Psychological distress cancer

• National Institute for Clinical Excellence
  (NICE) Guidance on Cancer Services (2004)Key
  recommendation 9
• Commissioners and providers of cancer services,
  working through Cancer Networks, should ensure
  that all cancer patients undergo systematic
  psychological assessment at key points and have
  access to appropriate psychological support.

6
Rasch Analysis of the HADS (1)

• Hospital Anxiety Depression Scale (Zigmond
  Snaith, 1983)
• Commonly used in health research / cancer
  (Herrmann, 1997)
• 14 items rated on a scale of 0 to 3
• I feel tense or wound up
• Worrying thoughts go through my mind
• I get sudden feelings of panic
• HADS-Anxiety(7) HADS-Depression(7) HADS-Total
  (14)
• Scores gt 14 HADS subscales definite case
• Scores gt 19 HADS are possibly cases

7
Rasch Analysis of the HADS (2)

• Screening efficacy of the HADS in cancer
  population
• Mixed results
• Some studies Good levels of sensitivity /
  specificity
• HADS Major depression / adjustment disorders
  (Razavi et al., 1990)
• HADS-D Depression (Berard et al., 1998 Katz et
  al., 2004)
• Others poor screening efficacy
• HADS, and subscales
• Hall et al., 1998
• Lloyd-Williams et al., 2001
• Love et al., 2001, 2002
• Despite conflicting results HADS remains widely
  used instrument in oncology

8
Rasch Analysis of the HADS (3)

• Issues
• Dimensionality
• 1, 2, 3, or 4 factor structure?
• Focus of questions
• Floor / ceiling effects?
• Would Rasch help?

9
Rasch Analysis of the HADS (4)

• Aim of study
• Explore the dimensionality of HADS using Rasch
• Identify misfitting items
• Evaluate screening efficacy misfitting items

10
Rasch Analysis of the HADS (5)

• Methods / Participants
• 1500 cancer patients
• Completed HADS on touchscreen
• Heterogeneous cancer diagnoses
• Subset of patients (n 382)
• Completed PSE / SCAN interview within 14 days of
  HADS
• Gold standard psychiatric interviews
• Diagnosis of psychological distress (case)

11
Rasch Analysis of the HADS (6)

• Methods / Participants
• Rasch Model
• Rating scale model (Andrich, 1978)
• Winsteps
• HADS, HADS-A, HADS-D

12
Rasch Analysis of the HADS (7)

• Methods / Participants
• Undimensionality
• Evaluated Item fit statistics
• Infit / Outfit mnsq range 0.70 1.3 (Wright et
  al., 1994)
• Principal components analysis
• Eigenvalues lt 3 (Linacre, 2005)
• Differential item functioning (male vs female,
  young (lt 55) vs older (gt 55)
• Item / person plots
• ROC curves of HADS vs modified HADS

13
Participants
14
Rasch Analysis of the HADS (8)

• Results
• Dimensionality
• Misfit (gt1.3) 1 item from HADS-A and HADS-D, 3
  items HADS-T
• PCA HADS (2.4), HADS-A and HADS-D (1.5)
• No DIF observed
• No differences gt than 0.5 logits observed (Bode
  et al., 2003)
• T-values lt 1.98

15
(No Transcript)
16
Rasch Analysis of the HADS (9)

• Results
• Item / Person plots

17
(No Transcript)
18
Rasch Analysis of the HADS (10)

• Results
• ROC curves of HADS vs modified HADS
• Removal of misfitting items had only marginal
  effect on sensitivity / specificity of HADS and
  subscales
• Range 0.67 0.70 (full / modified)

19
(No Transcript)
20
Rasch Analysis of the HADS (11)

• Conclusions
• HADS and subscales unidimensional structures
• Three items identified as misfitting
• Removal of misfitting items
• Screening efficacy modest
• However, items removed without detriment to
  screening
• Patient burden
• Little overlap between item locations and person
  measures
• Screening instrument
• Low / moderate levels of psychological distress
• Differences observed between studies -gt DIF
• Heterogeneous cancer populations vs single
  diagnosis groups

21
Acknowledgements

• Robert Rush - Centre for Integrated Health Care
  Research, Queen Margaret University College,
  Edinburgh
• CR UK Psychological Medicine Group, Western
  General Hospital, Edinburgh
• Smith, A.B., Wright, E.P., Rush, R., Stark, D.,
  Velikova, G. Selby, P.J. (in press). Rasch
  Analysis of the dimensional structure of the
  HADS. Psycho-Oncology

22
Development of an Item Bank for Screening
Assessing Psychological Distress in Cancer
Patients

• Cancer Research UK Psychosocial Clinical
  Practice Group, Leeds
• Cancer Research UK Psychological Medicine
  Group, Edinburgh

23
Development of an Item Bank Background

• Psychological distress in cancer patients
• Prevalence
• Anxiety disorders 7 - 23 (Stark House, 2000)
• Depression 7 - 47 (Sellick Crooks, 1999)
• Existing measures
• Disappointing efficacy (Cull et al., 2001)
• Large number of questions
• Lack of relevance
• Floor ceiling effects
• No consensus about which measures are optimal for
  which patient group (Holland et al., 2001)
• Difficulty comparing interventions assessed with
  different outcome measures

24
Development of an Item Bank Data

• Data collated from published (Velikova et al.,
  1999 Smith et al., 2002 Sharpe et al., 2004)
  unpublished studies conducted at
• CR UK Psychosocial Clinical Practice Group,
  St. Jamess University Hospital, Leeds
• CR UK Psychological Medicine Group, Western
  General Hospital, Edinburgh.
• 4910 cancer patients
• 83 items from 8 (touchscreen) questionnaires

25
Development of an Item Bank Items

• Hospital Anxiety Depression Scale (HADS,
  Zigmond Snaith, 1983)
• State-Trait Anxiety Inventory (STAI, Spielberger,
  1983)
• Beck Depression Inventory (BDI, Beck et al.,
  1996)
• Mental Health Inventory (MHI, Veit Ware, 1983)
• Patient Health Questionnaire (PHQ9, Kroenke
  Spitzer, 2002)
• General Health Questionnaire-12 (GHQ12, Goldberg
  Williams, 1991)
• Emotional Well-being scale of FACT (EWB, Cell et
  al., 1993)
• Emotional Functioning Scale of EORTC QLQ-c30
  (Aaronson et al., 1993)

26
Development of an Item Bank Items

• Rating scale model (Andrich, 1978)
• Common item equating
• Common items shared by each respondent
• Initial calibration of common item parameters
• Common item parameters used as anchors (held
  constant)
• Additional items added iteratively
• Item fit determined after each iteration
• Misfitting items removed
• Process repeated until all items have been added

27
(No Transcript)
28
Development of an Item Bank Results (1)

• Item fit
• 20 items removed due to misfit
• Remaining (63) items
• Cronbachs alpha 0.84
• Principal Components Analysis of residuals
• Unidimensional scale
• Good item fit (fit statistics lt 1.3)
• Items indicate non-specific distress

29
Items removed
Table 1 Removed items
30
Development of an Item Bank Results (2)

• Item person locations
• Items
• Face validity
• I.e. questions concerning suicidal ideation
  (BDI) or worthlessness are harder to endorse
  than inability to work or fatigue
• Mean item location -0.50 logits (s.e. 0.07,
  range -5 to 3)
• Persons
• Mean person location -2.21 logits (s.e. 1.01,
  range -5 to 1)
• Insufficient overlap between between items and
  persons
• Items targeting higher levels of distress

31
Final Item and Person Locations
32
Items retained
33
Development of an Item Bank Conclusions

• Rasch analysis
• Items pooled from a number of questionnaires
• Unidimensional construct for assessing
  psychological distress in cancer patients
• Individual items
• BDI covers entire range of psychological distress
• Other measures, such as STAI and EWB measure
  narrower range
• Additional items may be required, particularly at
  lower levels of distress

34
Development of an Item Bank Future Work

• Methodological work
• Rasch parameters sample size
• Exploration of measures
• Individual analysis
• GHQ12, PHQ9, STAI and BDI
• Analysis of Fit
• Data collection
• Additional measures
• Development of Computer-adaptive Test (CAT)
• Field work
• Comparison of CAT with
• Static questionnaire / Psychiatric interview
  (e.g. PSE / SCAN)