The GHQ-28 in one page: four subscales, two scoring methods, the Spanish cut-offs with their accuracy, and where to buy this commercial questionnaire.
Screens for non-psychotic psychological distress in four areas —somatic symptoms, anxiety and insomnia, social dysfunction and severe depression— by comparing recent state with usual functioning. Commercially licensed, with a Spanish validation that set its own cut-offs.
By Víctor Ciudad-Fernández, PhD · Updated: October 6, 2026 · Leer esta ficha en español · Instruments
The GHQ-28 is a screening tool, not a diagnostic one: it identifies people with a high probability of a non-psychotic psychiatric disorder, mainly anxiety and depression, by comparing their recent state with their usual functioning rather than asking about symptoms out of context. That “change from your usual self” logic is the hallmark of the whole GHQ family.
The 28-item version came from a factor analysis of the symptoms of the longer GHQ in a primary care setting, and organizes items into four subscales: somatic symptoms, anxiety and insomnia, social dysfunction and severe depression. Besides an overall screening score, it gives a profile by area that suggests which kind of distress predominates.
It is used mainly in primary care and in epidemiological and occupational health studies, to identify who needs a fuller clinical assessment, and as an outcome measure when research needs a brief indicator of general psychological distress rather than a categorical diagnosis.
Each item asks whether, in recent weeks, the person has noticed a symptom or a difficulty in functioning more, the same or less than usual, with four response options. Two scoring methods are in common use: the GHQ method (0-0-1-1), designed to classify probable “cases” against a cut-off, with a 0 to 28 range; and Likert scoring (0-1-2-3), which treats distress as a continuum, with a 0 to 84 range.
Cut-offs depend on the scoring method and the population: the Spanish ones below use the GHQ method.
Goldberg and Hillier (1979) derived the four subscales from the factor structure of GHQ symptoms in primary care, and found that structure to be very similar across three independent data sets. Those are the four subscales used to score the GHQ-28.
The subscales are not independent of each other: the Spanish validation notes that they are “by no means independent”, although their concurrent validity with clinical interview ratings suggests that they add information about anxiety and depression.
The Spanish validation reported two cut-offs with the GHQ method, one more specific and one more sensitive:
| Cut-off (GHQ method) | Accuracy | Source |
|---|---|---|
| 6/7 | 85% correctly classified; sensitivity 76.9%, specificity 90.2% | Lobo et al. (1986) |
| 5/6 | 83% correctly classified; sensitivity 84.6%, specificity 82% | Lobo et al. (1986) |
These cut-offs come from internal medicine outpatients. In another population, check them against a clinical criterion before using them to classify cases.
A screening questionnaire is not a diagnosis: a high score means a clinical assessment is warranted, not that a disorder is present.
Lobo, Pérez-Echeverría and Artal (1986) confirmed the validity of a Spanish version of the GHQ-28 in 100 internal medicine outpatients, assessed independently by psychiatrists with the Clinical Interview Schedule. Its ability to discriminate cases was almost as good as that of the Spanish GHQ-60, with the advantage of being much shorter.
License: Commercially licensed: there is no legitimate open-access version to link to. GL Assessment distributes the four official versions (GHQ-12, GHQ-28, GHQ-30 and GHQ-60).
Buy the questionnaire and the user’s guide from the publisher. For the terms of use of a specific translation, such as the Spanish adaptation by Lobo et al. (1986), contact the publisher or the authors of the validation.
We do not publish any item or copy of the questionnaire.
Cite the paper that introduced the scaled version and the language version you use:
@article{goldberg1979ghq28,
author = {Goldberg, D. P. and Hillier, V. F.},
title = {A scaled version of the {General Health Questionnaire}},
journal = {Psychological Medicine},
year = {1979},
volume = {9},
number = {1},
pages = {139--145},
doi = {10.1017/S0033291700021644}
}The Spanish validation is, at heart, an agreement problem between two ways of classifying the same person: the GHQ-28 cut-off and the diagnosis of a trained clinical interviewer. If your study compares the GHQ “case / non-case” classification with another criterion, the Krippendorff’s alpha calculator gives you the right categorical agreement index, with any number of raters and tolerating missing values.
The GHQ-12 is the short screener for surveys and gives a single score. The GHQ-28 adds four subscale scores (somatic symptoms, anxiety and insomnia, social dysfunction and severe depression), useful if you need a profile and not just a screen.
The publisher does not recommend any GHQ version for children, although it notes that several researchers have used it successfully with adolescents. If your sample is young, justify its use with studies in that age group.
This fact sheet is an academic reference for information purposes: it does not reproduce any questionnaire item. Using the instrument in your research or practice may require citing its authors or asking for permission: see “How to get the instrument”.