General Health Questionnaire – 28 (GHQ-28)

Evidence Reviewed as of before: 18-01-2015
Author(s)*: Annabel McDermott, OT
Editor(s): Annie Rochette, PhD OT
Content consistency: Gabriel Plumier

Purpose

The General Health Questionnaire – 28 (GHQ-28) is a self-report questionnaire that is used as a screening tool for psychological wellbeing.

In-Depth Review

Purpose of the measure

The General Health Questionnaire – 28 (GHQ-28) is self-report screening measure used to detect possible psychological disorder. The GHQ-28 identifies two main concerns: (1) the inability to carry out normal functions; and (2) the appearance of new and distressing phenomena (Goldberg & Hillier, 1979).

Available versions

The GHQ-28 is derived from the original 60-item General Health Questionnaire. There is also a 30-item version (GHQ-30) and a 12-item version (GHQ-12).

Features of the measure

Items:
The GHQ-28 consists of 28 questions designed to identify whether an individual’s current mental state differs from his/her typical state. Questions include:

Have you recently been feeling perfectly well and in good health?
Have you recently lost much sleep over worry?
Have you recently been managing to keep yourself busy and occupied?
Have you recently felt constantly under strain?
Have you recently felt that life is entirely hopeless?

Factor analysis of the GHQ-28 identified four 7-item subscales:
Somatic symptoms (items 1-7)
Anxiety/insomnia (items 8-14)
Social dysfunction (items 15-21)
Severe depression (items 22-28).

There is a high correlation between the anxiety subscale and the total score, showing that anxiety is a common symptom of psychiatric disorders (Goldberg & Hillier, 1979). Accordingly, subscales are not independent of each other and subscores should not be used to indicate specific psychological diagnoses. Rather, the measure is used to identify the presence of symptoms compared to what is normal for the individual (Salter et al., 2013).

Scoring:
The individual is asked to rate how he/she feels in relation to each question, according to the following criteria:

  • Better than usual
  • Same as usual
  • Worse than usual
  • Much worse than usual

Different scoring methods have been reported. One scoring method adopts a Likert scale of 0 to 3, resulting in a total possible score range of 0 to 84. This Likert scoring system was used with the original 60-item GHQ (Goldberg & Hillier, 1979).

An alternative and more common method attributes a binary score system of 0 to the first and second response options (better than usual, same as usual) and a score of 1 to the third and fourth response options (worse than usual, much worse than usual).

Some note that this scoring system is not sensitive to individuals with chronic conditions, where the individual may have experienced a symptom for a prolonged period of time (O’Rourke et al., 1998). Accordingly, the chronic scoring method attributes a score of 0 to the first item (better than usual) and a score of 1 to the third and fourth items, as per the traditional scoring method. The second item (‘same as usual’) receives a score of 0 for negative items and a score of 1 for positive items.

Response option Traditional (acute) scoring method Chronic scoring method Likert scoring method
Better than usual 0 0 (all items) 0
Same as usual 0 0 (negative items)1 (positive items) 1
Worse than usual 1 1 2
Much worse than usual 1 1 3

Higher scores indicate a greater possibility of psychological distress. A score ≥5 has been reported to indicate probable cases of psychiatric disorder (Anderson et al., 1996), however this has not been validated as the most appropriate score for the stroke population (Salter et al., 2013).

What to consider before beginning:
The choice of scoring method may impact diagnosis.
The GHQ-28 is not designed to detect chronic mental health conditions.

Time:
The GHQ-28 takes approximately 5 minutes to administer.

Training requirements
No training requirements have been specified for the GHQ-28, however it is advised that clinicians read the assessment manual prior to use.

Equipment
The GHQ-28 is a self-report questionnaire that does not require specific equipment.

Client suitability

Can be used with:

  • Individuals with stroke
  • Individuals with cardiac conditions
  • Individuals with spinal cord injury
  • Individuals with musculoskeletal conditions
  • The elderly (Rehabilitation Measures Database, 2010)

Should not be used with:

  • The GHQ-28 has not been reported to be unsuitable for use with any particular population.

In what languages is the measure available?

  • The GHQ-28 is available in 38 languages (Sterling, 2011) and has cross-cultural applicability (Kilic et al., 1997).

Summary

What does the tool measure? Psychological wellbeing.
What types of clients can the tool be used for? The GHQ-28 can be used with, but is not limited to, patients with stroke.
Is this a screening or assessment tool? Screening.
Time to administer Five minutes.
ICF Domain Body Function.
Versions GHQ (original 60-item version)
GHQ-30
GHQ-28
GHQ-12
Other Languages The GHQ is available in 38 languages
Measurement Properties
Reliability Internal consistency:
No studies have reported on internal consistency of the GHQ in a sample of individuals with stroke.

Test-retest:
One study reported excellent test-retest reliability of the GHQ-28 in a sample of individuals with stroke (time since stroke not specified).

Intra-rater:
No studies have reported on the intra-rater reliability of the GHQ in a sample of individuals with stroke.

Inter-rater:
No studies have reported on the inter-rater reliability of the GHQ in a sample of individuals with stroke.

Validity Criterion:
Concurrent:
– One study reported excellent concurrent validity between the GHQ-28 total score and the Zung Self-Rating Depression Scale, Hamilton Depression Scale and the Present State Examination.
– One study reported no difference between the GHQ-30 and HAD Scale total scores when identifying any DSM-IV diagnosis, anxiety or depression.

Predictive:
One study reported that patients with acute to chronic stroke who were diagnosed as depressed according to ICD-10 or DSM-IIIR criteria achieved a significantly higher GHQ-28 total score than patients who were not diagnosed as depressed; a score >4 on the GHQ-28 correlated with depression among participants.

Construct:
Convergent/Discriminant:
– One study reported adequate correlations between the GHQ-12 and the Stroke and Aphasia Quality of Life scale (SAQOL) and SAQOL-39 mean scores; adequate correlations between the GHQ-12 and SAQOL subtests; and adequate to excellent correlations between the GHQ-12 and SAQOL-39 subtests.
– One study reported that individuals with stroke with a Beck Depression Inventory (BDI) score of 11-18 (mild depression) demonstrated GHQ-28 median scores of 27.0 and 28.0 at 1 and 6 months post-stroke respectively; individuals with a BDI score ≥19 (severe depression) demonstrated GHQ-28 median scores of 44.0 and 48.0 at 1 and 6 months post-stroke respectively.
– No studies have reported on cross-diagnostic validity of the GHQ in a sample of individuals with stroke.

Known Groups:
No studies have reported on known-group validity of the GHQ in a sample of patients with stroke.

Floor/Ceiling Effects No studies have reported on floor/ceiling effects of the GHQ within a sample of individuals with stroke.
Does the tool detect change in patients? The GHQ-28 is intended for use as a screening instrument and therefore is not designed to measure change over time.
– One study reported 81% sensitivity and 68% specificity of the GHQ-28 when using cutoff scores of 11/12 (optimal in relation to DSM-IIIR criteria), or 85% sensitivity and 61% specificity when using cutoff scores of 7/8 (optimal in relation to ICD-10 criteria).
– One study examined reported 80% sensitivity and 76% specificity of the GHQ-30 when using a cutoff score of 8/9.
Acceptability The GHQ-28 is non-invasive and quick to administer. Caution should be exercised with scoring.
Feasibility The GHQ-28 is suitable for administration in various settings. The assessment is quick to administer and requires minimal specialist equipment or training.
How to obtain the tool?

https://www.gl-assessment.co.uk/products/general-health-questionnaire-ghq/

Psychometric Properties

Overview

A literature search was conducted to identify all relevant publications on the psychometric properties of the GHQ relevant to the stroke population. Five studies were identified. Please note that three of these studies use the GHQ-28 (Lincoln et al., 2003; Robinson & Price, 1982; Thomas & Lincoln, 2006); one study that used the GHQ-30 (O’Rourke et al., 1998) and one study that used the GHQ-12 (Hilari et al., 2003) were also included.

Floor and ceiling effect

No studies have reported on the floor or ceiling effects of the GHQ in a sample of individuals with stroke.

Reliability

Internal consistency:
No studies have reported on internal consistency of the GHQ in a sample of individuals with stroke.

Test-retest:
Robinson and Price (1982) examined test-retest reliability of the GHQ-28 with a sample of 20 individuals (time since stroke not specified) and reported excellent 2-month test-retest reliability (r=0.90).

Intra-rater:
No studies have reported on the intra-rater reliability of the GHQ in a sample of individuals with stroke.

Inter-rater:
No studies have reported on the inter-rater reliability of the GHQ in a sample of individuals with stroke.

Validity

Content:

The GHQ-28 is a scaled version of the original 60-item GHQ developed by Goldberg in 1978. Factor analysis of the original GHQ was conducted in a sample of 523 individuals who attended a primary care setting, resulting in the 28-item version with four 7-item subscales (Goldberg & Hillier, 1979).

Criterion:

Concurrent:
Robinson and Price (1982) examined concurrent validity of the GHQ-28 in a sample of 103 individuals with stroke (time since stroke not specific) by comparison with other psychopathology scales. The authors reported excellent concurrent validity between the GHQ-28 total score and the Zung Self-Rating Depression Scale (r=0.86), Hamilton Depression Scale (r=0.88) and the Present State Examination (r=0.94).

O’Rourke et al. (1998) examined concurrent validity of the GHQ-30 in a sample of 105 individuals with chronic stroke by comparison with the Hospital Anxiety and Depression (HAD) Scale. There was no difference between the GHQ-30 and HAD Scale total scores when identifying any DSM-IV diagnosis (p=0.95), anxiety (p=0.25) or depression (p=0.56), using ROC curves.
Note: The study used the conventional 0-0-1-1 format to score the GHQ-30; this version of the GHQ is not split into subscales for depression and anxiety.

Predictive:
Lincoln et al. (2003) examined predictive validity of the GHQ-28 in a mixed sample of 143 stroke patients with acute to chronic stroke. Patients who were diagnosed as depressed according to the ICD-10 or DSM-IIIR achieved a significantly higher (p≤0.01) GHQ-28 total score than patients who were not diagnosed as depressed (ICD-10: kappa=0.40, IQR depressed 9-19/not depressed 3-12; DSM-IIIR: kappa=0.12, IQR depressed 12-21/not depressed 5-13). A score >4 on the GHQ-28 correlated with depression among participants.

GHQ-28 ICD-10 DSM-IIIR
Depressed (42%) Not depressed (52%) Depressed (15%) Not depressed (77%)
IQR 9-19 3-12 12-21 5-13
Kappa 0.40 0.12

Construct:

Convergent/Discriminant :
Hilari et al. (2003) examined convergent validity of the GHQ-12 in a sample of 83 individuals with chronic stroke and aphasia, by comparison with the Stroke and Aphasia Quality of Life scale (SAQOL) and the SAQOL-39. The study yielded an adequate correlation between the GHQ-12 and SAQOL mean (r=0.58, p<0.01) and between the GHQ-12 and the SAQOL-39 mean (0.53, p<0.01). Correlations between the GHQ-12 and SAQOL subtests were adequate (mood r=0.57, thinking r=0.41, personality r=0.57, energy r=0.32, family roles r=0.41, social roles r=0.41, work r=0.34, p<0.01). Correlations between the GHQ-12 and SAQOL-39 subtests were adequate (physical r=0.39, energy r=0.32, p<0.01) to excellent (psychosocial r=0.62, p<0.01).

Thomas and Lincoln (2006) reported on convergent validity of the GHQ-28 in a sample of 123 individuals with stroke and depression, by comparison with the Beck Depression Inventory (BDI). Measures were taken at 1 month and 6 months post-stroke. Individuals who were diagnosed with mild depression (BDI score of 11-18) demonstrated GHQ-28 median scores of 27.0 (IQR=21.5-36.0) and 28.0 (IQR=22.0-37.0) at 1 and 6 months post-stroke respectively. Individuals with severe depression (BDI score ≥19) demonstrated GHQ-28 median scores of 44.0 (IQR=32.0-54.5) and 48.0 (IQR=35.0-55.0) at 1 and 6 months post-stroke respectively.

Known Group:
No studies have reported on known-group validity of the GHQ in a sample of patients with stroke.

Responsiveness

Sensitivity & Specificity:
Lincoln et al. (2003) examined sensitivity and specificity of the GHQ-28 in a mixed sample of 143 stroke patients with acute to chronic stroke. The study found that optimal cutoff scores for the GHQ-28 in relation to DSM-IIIR and ICD-10 criteria were 11/12 (sensitivity 81%, specificity 68%) and 7/8 (sensitivity 85%, specificity 61%) respectively.

GHQ-28 cutoff score ICD-10 diagnosis DSM-IIIR diagnosis
Sensitivity Specificity Sensitivity Specificity
5 0.98 0.35 1.00 0.24
6 0.98 0.44 1.00 0.29
7 0.88 0.55 0.95 0.41
8 0.85 0.61 0.95 0.47
9 0.78 0.63 0.95 0.52
10 0.72 0.68 0.86 0.57
11 0.63 0.72 0.81 0.63
12 0.57 0.73 0.81 0.68
13 0.48 0.76 0.76 0.73
14 0.47 0.80 0.71 0.76
15 0.43 0.84 0.67 0.80

O’Rourke et al. (1998) examined sensitivity and specificity of the GHQ-30 in a sample of 105 individuals with chronic stroke. Using previously recommended cutoff scores of 4/5 yielded sensitivity and specificity scores of 0.9 and 0.47 (respectively). The authors recommended a cutoff score of 8/9, which achieved sensitivity and specificity scores of 0.8 and 0.76 (respectively).

References

Anderson, C., Laubscher, S., & Burns, R. (1996). Validation of the Short Form 36 (SF-36) health survey questionnaire among stroke patients. Stroke, 27, 1812-6.

Goldberg, D.P. & Hillier, V.F. (1979). A scaled version of the General Health Questionnaire. Psychological Medicine, 9, 139-45.

Hilari, K., Byng, S., Lamping, D.L., & Smith, S.C. (2003). Stroke and Aphasia Quality of Life Scale-39 (SAQOL-39): Evaluation of acceptability, reliability and validity. Stroke, 34, 1944-50.

Kilic, C., Rezaki, M., Rezaki, B., Kaplan, I., Ozgen, C., Sagduyu, A., & Ozturk, M.O. (1997). General Health Questionnaire (GHQ12 & GHQ28): psychometric properties and factor structure of the scales in a Turkish primary care sample. Social Psychiatry and Psychiatric Epidemiology, 32, 327-31.

Lincoln, N.B., Nicholl, C.R., Flannaghan, T., Leonard, M., & Van der Gucht, E. (2003). The validity of questionnaire measures for assessing depression after stroke. Clinical Rehabilitation, 17, 840-6.

Malakouti, S.M., Fatollahi, P., Mirabzadeh, A., & Zandi, T. (2007). Reliability, validity and factor structure of the GHQ-28 used among elderly Iranians. International Psychogeriatrics, 19(4), 623-34.

O’Rourke, S., MacHale, S., Signorini, D., & Dennis, M. (1998). Detecting psychiatric morbidity after stroke: Comparison of the GHQ and HAD Scale. Stroke, 29, 980-5.

Rehabiliation Measures Database. (2010). General Health Questionnaire-28. Retrieved from http://www.rehabmeasures.org/Lists/RehabMeasures/PrintView.aspx?ID=909

Robinson, R.G. & Price, T.R. (1982). Post-stroke depressive disorders: A follow-up study of 103 patients. Stroke, 13(5), 635-40.

Salter, K., Campbell, N., Richardson, M., Mehta, S., Jutai, J., Zettler, L., Moses, M., McClure, A., Mays, R., Foley, N., & Teasell, R. (2013). Outcome Measures in Stroke Rehabilitation. Retrieved from http://www.ebrsr.com/sites/default/files/Chapter21_Outcome-Measures_FINAL_16ed.pdf

Sterling, M. (2011). General Health Questionnaire – 28 (GHQ-28). Journal of Physiotherapy, 57, 259.

Thomas, S.A. & Lincoln, N.B. (2006). Factors relating to depression after stroke. British Journal of Clinical Psychology, 45, 49-61.

Werneke, U., Goldberg, D.P., Yalcin, I., & Ustun, B.T. (2000). The stability of the factor structure of the General Health Questionnaire. Psychological Medicine, 30, 823-9.

Willmott, S.A., Boardman, J.A.P., Henshaw, C.A., & Jones, P.W. (2004). Understanding General Health Questionnaire (GHQ-28) score and its threshold. Social Psychiatry and Psychiatric Epidemiology, 39, 613-7.

See the measure

How to obtain theGeneral Health Questionnaire – 28 (GHQ-28)?

https://www.gl-assessment.co.uk/products/general-health-questionnaire-ghq/

Table of contents
Survey