WHOQOL-BREF

WHOQOL-BREF in Trials, Surveys and Economic Evaluation

The WHOQOL-BREF is the short form of the World Health Organization's quality of life assessment, built to describe a whole life and not only health, in far fewer questions than its parent instrument, the WHOQOL-100. For a health economist it sits on the outcome side of an evaluation as a descriptive measure. Its four domain scores show where a programme changes people's lives, but they carry no preference weights, so they cannot be multiplied by time to give QALYs. Studies that want both a broad picture of quality of life and a cost per QALY therefore pair it with a preference-based measure, or use a published mapping algorithm, such as the one Kangwanrattanakul estimated in the general Thai population, to predict EQ-5D-5L index values from WHOQOL-BREF responses.

A definition of quality of life wider than health

The WHOQOL instruments start from a definition set out by the WHOQOL Group in its 1995 position paper. Quality of life is "individuals' perception of their position in life in the context of the culture and value systems in which they live and in relation to their goals, expectations, standards and concerns". The definition is subjective, and it places health inside a wider life, which is why the short form has social relationships and environment domains as well as physical and psychological ones.

That starting point separates the WHOQOL family from most measures of health-related quality of life. A health status questionnaire asks what a person can do and what symptoms the person has. The WHOQOL-BREF also asks how the person judges life in areas that health care touches only indirectly, which is why it has an environment domain. For an economist this breadth matters when an intervention works through housing, income, social contact or safety, because a narrow health measure may miss those effects.

The longer parent instrument is the WHOQOL-100. The WHOQOL Group described the WHOQOL-BREF in 1998 as an abbreviated version of it, derived from data collected with the WHOQOL-100, and the international field trial report by Skevington and colleagues describes the short form as a 26-item version of the WHOQOL-100.

Items, domains and the WHO scoring rule

The WHOQOL-BREF reports four domains, physical health, psychological, social relationships and environment. Alongside them sit questions on overall quality of life and general health, which the 1998 paper describes as one facet and which are scored apart from the four domains.

The WHO user manual for the WHOQOL-100 and WHOQOL-BREF sets the scoring. A domain score is the mean of the items in that domain multiplied by four, and the result is then transformed to a 0 to 100 scale.

$$D_k = 4 \times \frac{1}{n_k} \sum_{i \in k} x_i$$

where $D_k$ is the raw score for domain $k$, $n_k$ is the number of items in that domain, $x_i$ is the coded response to item $i$, and the sum runs over the items that belong to domain $k$. The manual then gives the conversion of $D_k$ to the 0 to 100 scale, on which a higher score means better quality of life.

The output is four numbers per person, plus the overall items. The WHO scoring reports the domains separately, and it attaches no weight to say how much a point of environment is worth against a point of physical health. That makes the WHOQOL-BREF a profile instrument, and the point shapes every economic use of it.

Evidence from development and the international field trial

The WHOQOL Group reported in 1998 that WHOQOL-BREF domain scores correlated at 0.89 or above with WHOQOL-100 domain scores calculated on the same four-domain structure. The short form also showed good discriminant validity, content validity, internal consistency and test-retest reliability. The authors judged it a valid and reliable alternative to the full instrument, best suited to studies that need a brief assessment, such as large epidemiological studies and clinical trials.

Skevington and colleagues then tested the instrument on cross-sectional survey data from adults in 23 countries, 11,830 people in all. The sample included sick and well respondents, drawn from the general population and from hospital, rehabilitation and primary care settings. Analyses of internal consistency, item-total correlations, discriminant validity and construct validity through confirmatory factor analysis led the authors to conclude in 2004 that the WHOQOL-BREF had good to excellent psychometric properties of reliability and performed well in preliminary tests of validity.

The multinational design supports use of the instrument in many countries. It does not by itself make mean scores comparable across countries, because any comparison still rests on how well each language version works.

Why the WHOQOL-BREF cannot supply QALY weights

A preference-based score, the kind that yields a health state utility value, has two properties the WHOQOL-BREF lacks. Its values come from people's stated preferences between health states, elicited with methods such as time trade-off or standard gamble, and the scale is anchored so that full health is 1 and dead is 0. Only a value with those properties can be multiplied by time to give QALYs.

The WHOQOL-BREF domain scores are means of rated items, rescaled. A score of 0 marks the bottom of the response range, not a state equivalent to death, and a score of 100 marks the top of the range, not full health in the QALY sense. The domains also have no exchange rate between them. Treating the 0 to 100 transformation as if it were a utility is a misuse that the scoring does not support.

The WHOQOL-BREF is therefore a non-preference-based generic measure. That classification describes how the scores were built and says nothing against the instrument's usefulness. The WHOQOL-BREF can record effects of a social or community programme that a narrower health measure misses, even though it cannot price those effects in QALYs.

Uses in trials and economic evaluation

The WHOQOL-BREF enters economic work as a secondary outcome beside a utility measure, as one of the outcomes in a cost-consequence analysis, and as source data for a mapping study. Each use treats the domain scores as description, never as values.

Secondary outcome alongside a utility measure

In a trial-based cost-utility analysis the QALY usually comes from a preference-based measure collected at the same visits, and the WHOQOL-BREF is reported as a secondary outcome. The pairing lets the analyst show whether the QALY result agrees with the broader quality of life picture. A gain in the social relationships or environment domains with no movement in the utility measure is a signal that the QALY may be missing part of the benefit.

Cost-consequence reporting

A cost-consequence analysis sets costs beside a list of outcomes without combining them. The four WHOQOL-BREF domains suit that format, because each can be shown with its own estimate and confidence interval and decision makers can weigh them openly. The price of this transparency is that no single ratio comes out, and two programmes with different domain profiles cannot be ranked without a judgement about which domains matter more.

Mapping to a utility scale

When a study collected the WHOQOL-BREF but no preference-based measure, a mapping function can predict utility values from WHOQOL-BREF responses. Kangwanrattanakul, writing in 2023, used data from 1,200 people in the general Thai population, randomly split into an estimation group and a validation group, and fitted nine regression models by curve estimation. An inverse relationship gave the best fit for predicting the EQ-5D-5L index score, and in 2023 the author called for the mapping equation to be investigated further in larger samples of the Thai population. A mapping estimated in one country and value set should not be assumed to hold for a different population or a different EQ-5D-5L value set.

Illustrative worked example with two community programmes

All numbers in this example are illustrative. A single three-arm trial compares two community programmes for adults with long-term mental health problems with usual care over 12 months. The trial collects the WHOQOL-BREF and the EQ-5D-5L in every arm.

1. Domain score for one respondent. A respondent's physical health items have a mean of 3.5. Under the WHO rule the raw domain score is 3.5 × 4 = 14, which the manual then converts to the 0 to 100 scale.

2. The two profiles. Mean differences from usual care on the 0 to 100 domain scales are as follows.

Programme A does more for psychological and social life, while Programme B does more for physical health. No rule in the WHOQOL-BREF says which profile is better.

3. Why averaging the domains does not solve the problem. A simple average gives (2 + 8 + 5 + 1) ÷ 4 = 4 points for A and (6 + 3 + 2 + 1) ÷ 4 = 3 points for B. That ranking assumes every domain point is worth the same, which is a value judgement that no respondent was asked to make.

4. Cost per QALY from the paired utility measure. Programme A costs GBP 1,200 more per person than usual care and its EQ-5D-5L utility is 0.03 higher for the year, so it gains 0.03 × 1 = 0.03 QALYs.

$$\text{ICER}_A = \frac{1200}{0.03} = 40000$$

where 1200 is the incremental cost in GBP, 0.03 is the incremental QALY gain and the result is the ICER in GBP per QALY. Programme B costs GBP 900 more and gains 0.03 QALYs, so its ICER is 900 ÷ 0.03 = GBP 30,000 per QALY.

5. Interpretation. Programme B gains the same QALYs as Programme A at a lower cost, so on the EQ-5D-5L evidence B dominates A. The WHOQOL-BREF profile shows that Programme A delivers its benefit mainly in the psychological and social domains, which a utility measure built around health may weight lightly. A decision maker who sees only the ICERs misses that difference, and one who sees only the profiles has no way to set the gains against the costs.

How the WHOQOL-BREF differs from neighbouring measures

Several measures sit close to the WHOQOL-BREF, and the differences change what a study can claim. The WHOQOL-BREF is a generic instrument, meant for use across conditions, in contrast with a disease-specific instrument that targets the symptoms of one condition. It is also a patient-reported outcome measure, because respondents answer about their own lives, although that label says nothing about whether scores are preference-based.

The SF-36 and the SF-12 are generic health status profiles centred on functioning and symptoms, and Brazier, Roberts and Deverill estimated a preference-based measure, the SF-6D, from the SF-36. The WHOQOL-BREF reaches further into a person's environment and social life, and has no equivalent preference-based scoring. The Nottingham Health Profile is an older generic profile whose statements describe health problems and distress, while the WHOQOL-BREF covers quality of life more widely, including social relationships and environment.

Quality of life, the construct the WHOQOL-BREF measures, is broader than health-related quality of life, the part that health and health care affect. A trial that reports the WHOQOL-BREF as its quality of life outcome is therefore measuring something wider than the HRQoL captured by a utility instrument, and the two need not move together.

Limitations and common misreadings

The instrument is well validated as a descriptive measure. Most problems arise when its scores are asked to do an economic job they were not built for.

• Domain scores read as utilities. A 0 to 100 domain score is not a utility, and dividing it by 100 does not make it one.
• Adding or averaging the domains. Any single total from the four domains imposes equal or arbitrary weights that no valuation study supports.
• Assuming a mapping travels. The Thai mapping was estimated in one population against one EQ-5D-5L value set, and its author asked for further testing. Using it elsewhere needs a check of fit in the new data.
• Overlooking breadth in the QALY. A study that reports only the QALY from a paired utility measure can understate programmes whose benefit lies in social or environmental domains. The WHOQOL-BREF results should be reported, not discarded once the ICER is known.
• Cross-country comparison. The international field trial supports the instrument's psychometric properties across settings, but comparing mean scores between countries still depends on each language version.

Sources

• Kangwanrattanakul K. Mapping of the World Health Organization Quality of Life Brief (WHOQOL-BREF) to the EQ-5D-5L in the general Thai population. PharmacoEconomics Open. 2023;7(1):139-148. doi:10.1007/s41669-022-00380-0
• Skevington SM, Lotfy M, O'Connell KA. The World Health Organization's WHOQOL-BREF quality of life assessment: psychometric properties and results of the international field trial. A report from the WHOQOL group. Quality of Life Research. 2004;13(2):299-310. doi:10.1023/B:QURE.0000018486.91360.00
• The WHOQOL Group. The World Health Organization quality of life assessment (WHOQOL): position paper from the World Health Organization. Social Science and Medicine. 1995;41(10):1403-1409. doi:10.1016/0277-9536(95)00112-K
• The WHOQOL Group. Development of the World Health Organization WHOQOL-BREF quality of life assessment. Psychological Medicine. 1998;28(3):551-558. doi:10.1017/S0033291798006667
• World Health Organization, Programme on Mental Health. WHOQOL user manual. WHO/MNH/MHP/98.4.Rev.1. Geneva, World Health Organization; 1998, revised 2012 (WHO/HIS/HSI Rev.2012.03). https://www.who.int/publications/i/item/WHO-HIS-HSI-Rev.2012-3
• Brazier J, Roberts J, Deverill M. The estimation of a preference-based measure of health from the SF-36. Journal of Health Economics. 2002;21(2):271-292. doi:10.1016/S0167-6296(01)00130-8
• Drummond MF, Sculpher MJ, Claxton K, Stoddart GL, Torrance GW. Methods for the Economic Evaluation of Health Care Programmes. 4th ed. Oxford University Press; 2015.