WEMWBS

WEMWBS as a Mental Wellbeing Outcome in Economic Evaluation

The Warwick-Edinburgh Mental Wellbeing Scale asks people only about positive aspects of mental health, such as feeling good about themselves or feeling cheerful, and adds the answers into a single total between 14 and 70. Health economists meet it in evaluations of public mental health, workplace, arts and community programmes, where many participants report no problems on a health-related quality of life measure such as the EQ-5D and a wellbeing outcome can still move. The total is a summed rating scale and not a utility, so it cannot be multiplied by time to give QALYs, and much of the economic interest in the scale concerns what can and cannot be done with a WEMWBS result in a cost-utility analysis. The seven-item short form, with its own Rasch-based scoring and a UK value set, is a separate measure, SWEMWBS.

Fourteen positively worded items and a summed score

The scale was built to monitor mental wellbeing in the general population and to evaluate projects and policies that aim to improve it. Tennant and colleagues reported its development in 2007. An expert panel drew on the academic literature, qualitative work with focus groups and psychometric testing of an existing scale.

Stewart-Brown and colleagues described the result as a measure of mental well-being focusing entirely on positive aspects of mental health, made up of 14 positively phrased Likert-style items. None of the items asks about symptoms or distress, which separates WEMWBS from depression and anxiety scales used in the same trials.

The University of Warwick's scoring guidance scores each item from 1 to 5 and sums the 14 item scores, so the total runs from 14 to 70. Higher totals mean better mental wellbeing. Because every item is positively worded, none is reverse scored, and no weights are applied, so every item counts equally and every step between response options is treated as the same size.

Validation in students, the general population and teenagers

The original validation used a student sample and a representative population sample in the UK. Confirmatory factor analysis supported a single underlying factor, which justifies adding the items into one total.

Cronbach's alpha was 0.89 in the student sample and 0.91 in the population sample. Tennant and colleagues read these values as suggesting some item redundancy in the scale. The scale showed no ceiling effect in the population sample and its distribution was near normal, a useful property for programmes whose participants start in reasonable health. Maheswaran and colleagues later cited a one-week test-retest reliability of 0.83 from the adult validation.

Clarke and colleagues extended the validation to teenage school students in England and Scotland. In 2011 they reported that 1,650 teenagers completed the questionnaire, that confirmatory factor analysis again showed one underlying factor, and that the test-retest intraclass correlation was 0.66 in 212 students. Students in focus groups generally found the scale comprehensible and acceptable, although some items were open to misinterpretation.

Population reference values from the Scottish Health Survey

The Scottish Health Survey carries WEMWBS for adults, which gives evaluators a national reference against which to read a programme's baseline. Its reports show adult mean scores between 49.4 and 50.0 over 2008 to 2019.

The mean then fell to 47.0 in 2022, outside the pre-pandemic range, rose to 48.9 in 2023 and was 48.5 in 2024. In 2024 the highest mean was among people aged 65 to 74 (50.4). National means have therefore varied by about three points over the period, and a trial sample that starts several points below the national mean is a group with lower wellbeing than the general adult population.

Responsiveness and the size of an important change

An outcome used in an economic evaluation has to detect change when change happens. Maheswaran and colleagues assessed the responsiveness of WEMWBS in 2012 with data from twelve intervention studies.

At group level the standardised response means ranged from minus 0.10 to 1.35, so some interventions moved the scale a great deal and others hardly at all. At individual level the standard error of measurement ranged from 2.4 to 3.1 points, and with a threshold of 2.77 standard errors of measurement the scale detected important improvement in at least 12.8% to 45.7% of participants in the studies. The authors concluded that WEMWBS is responsive to changes occurring in a wide range of mental health interventions.

These thresholds are distribution based. They describe change larger than measurement error, which is a different idea from a minimal clinically important difference anchored on what participants regard as a worthwhile improvement.

Why a WEMWBS total cannot be turned into QALYs directly

WEMWBS is a non-preference-based measure. Its scoring is a simple sum that draws on no valuation of WEMWBS states, so the total has no anchor at full health or dead and none of the interval meaning that a QALY calculation needs.

The obvious alternative, predicting EQ-5D utilities from WEMWBS with a mapping function, depends on the two measures sharing enough content. Johnson and colleagues tested this in 2016 with the Coventry Household Survey, in which 7,469 participants completed the EQ-5D-3L, WEMWBS and a self-rated health question. About 75% of participants scored the maximum on the EQ-5D-3L. The relationship was weak, with an R squared of 0.104 for the EQ-5D-3L index and 0.141 for the visual analogue scale, and levels of mental wellbeing varied greatly among participants who had the maximum EQ-5D-3L score. The authors concluded that a new measure, a Wellbeing Adjusted Life Year, is needed to evaluate the relative effectiveness of interventions that affect mental wellbeing.

Later preference work turned to the short form. A UK value set for SWEMWBS and a mapping from SWEMWBS to the ReQoL utility index now exist, and both use the seven short-form items, which a WEMWBS administration contains. A mental wellbeing adjusted life year built on that value set is a different unit from a QALY based on the EQ-5D.

Use in programme evaluations and in NICE guidance

The scale has been used in a large body of intervention research outside clinical medicine. Blodgett and colleagues reviewed 223 interventions in 209 studies that measured wellbeing with the 7 or 14-item scale before and after the intervention, covering about 53,834 participants.

Psychological approaches based on resilience, wellbeing or self-management had the strongest effects, and person-centred advice, arts-based interventions, parenting programmes and social prescribing showed medium to strong effects. Effects were pooled as standardised mean differences within the intervention groups, comparing scores before and after each intervention. Within-group before-and-after changes in WEMWBS cannot be treated as treatment effects in a cost-effectiveness model, because wellbeing scores can drift up or down for reasons unrelated to the programme.

NICE names the scale in public health guidance. NICE guideline NG212 on mental wellbeing at work (published March 2022) says in recommendation 1.1.7 that a validated measure should be used when measuring mental wellbeing, and gives Warwick Medical School's Warwick-Edinburgh Mental Wellbeing Scales as one example alongside the government's voluntary reporting framework for large employers and the What Works Wellbeing workplace questionnaire. The recommendation concerns measurement by employers.

Where an economic evaluation of such a programme uses WEMWBS, the result can be reported in a cost-consequence analysis or as a cost per point gained, beside any EQ-5D based QALYs.

Illustrative worked example of a cost per WEMWBS point

All figures here are illustrative. A community wellbeing programme is evaluated against usual activities in a controlled study. Mean WEMWBS scores rise by 4.0 points in the programme group and by 1.5 points in the control group over six months. The programme adds GBP 150 per participant to costs.

1. Incremental effect. The difference in mean change is 4.0 − 1.5 = 2.5 WEMWBS points.

2. Cost per point gained.

$$\text{Cost per point} = \frac{\Delta C}{\Delta W_I-\Delta W_C}$$

where $\Delta C$ is the incremental cost per participant in GBP, $\Delta W_I$ is the mean change in WEMWBS in the programme group and $\Delta W_C$ is the mean change in the control group.

$$\frac{150}{4.0-1.5} = \frac{150}{2.5} = 60$$

where 150 is the incremental cost in GBP and 2.5 is the incremental WEMWBS change, giving GBP 60 per point gained.

3. Cohort cost. For 200 participants the programme costs 200 × 150 = 30,000, or GBP 30,000, for a total of 200 × 2.5 = 500 extra WEMWBS points.

4. Individual change. Taking an illustrative standard error of measurement of 2.8 points, inside the range Maheswaran and colleagues reported, the 2.77 threshold is 2.77 × 2.8 = 7.756 points. A participant would need to improve by about 8 points to count as an individual improvement under that rule, which is far larger than the mean group effect of 2.5.

GBP 60 per point has no external benchmark. The cost per point cannot be set against a cost-per-QALY threshold, because a WEMWBS point has no utility value, and it can be compared only with the cost per point of other programmes measured on the same scale. A decision maker seeking a cost per QALY would need EQ-5D data collected in the same study, or a preference-based route through the short form.

How WEMWBS differs from neighbouring measures

WEMWBS sits among several measures that a health economist might choose for a mental wellbeing outcome. The choice depends on whether the evaluation needs a profile of positive mental health, a utility for QALYs or a capability score.

SWEMWBS uses seven of the fourteen items and converts its raw total to a Rasch-based metric score. ICECAP-A measures capability wellbeing in adults. The EQ-HWB was developed to capture health and wellbeing for patients, social care users and carers in economic evaluation. ReQoL-10 belongs to the Recovering Quality of Life measures used in mental health. WEMWBS differs from all of these in measuring only positive mental wellbeing, with no utility or capability scoring of the full scale, and from symptom scales such as depression questionnaires in containing no items about distress. Wellbeing in the broad sense, covering comfort, health and happiness as a whole, is a wider construct than the mental wellbeing WEMWBS measures.

Limitations and common misreadings

Several errors recur when WEMWBS results reach an economic evaluation. Each changes how the evidence should be used.

• Treating the total as a utility. A WEMWBS total of 56 is not 0.8 of anything. Dividing by 70, or rescaling 14 to 70 onto 0 to 1, produces a number that looks like a utility but rests on no preference data.
• Assuming equal intervals. The summed score treats every item and every response step as equal. In the 2009 Rasch analysis, Stewart-Brown and colleagues found that the full 14 items fitted the model poorly and that three items showed significant misfit, and many of the poorly fitting items showed considerable bias by gender.
• Reading within-group change as effect. The 2022 review pooled changes before and after each intervention within the intervention groups, and such pre and post differences need a comparator before they enter a model.
• Confusing measurement error with importance. The standard error of measurement thresholds from 2012 describe change beyond noise. They are not an anchor-based minimal important difference.
• Mixing the long and short forms. A WEMWBS total and a SWEMWBS metric score are on different scales and cannot be pooled or compared directly.
• Overlooking licensing. The University of Warwick issues commercial and non-commercial licences for the scales, counted by the number of people offered the scale and valid for 12 months, and an evaluation plan has to allow for the licence.

Sources

• Blodgett JM, Birch JM, Musella M, Harkness F, Kaushal A. What works to improve wellbeing? A rapid systematic review of 223 interventions evaluated with the Warwick-Edinburgh Mental Well-Being Scales. International Journal of Environmental Research and Public Health. 2022;19(23):15845. doi:10.3390/ijerph192315845
• Clarke A, Friede T, Putz R, et al. Warwick-Edinburgh Mental Well-being Scale (WEMWBS): validated for teenage school students in England and Scotland. A mixed methods assessment. BMC Public Health. 2011;11:487. doi:10.1186/1471-2458-11-487
• Johnson R, Jenkinson D, Stinton C, Taylor-Phillips S, Madan J, Stewart-Brown S, Clarke A. Where's WALY? A proof of concept study of the 'wellbeing adjusted life year' using secondary analysis of cross-sectional survey data. Health and Quality of Life Outcomes. 2016;14:126. doi:10.1186/s12955-016-0532-5
• Maheswaran H, Weich S, Powell J, Stewart-Brown S. Evaluating the responsiveness of the Warwick Edinburgh Mental Well-Being Scale (WEMWBS): group and individual level analysis. Health and Quality of Life Outcomes. 2012;10:156. doi:10.1186/1477-7525-10-156
• National Institute for Health and Care Excellence. Mental wellbeing at work (NG212). Published 2 March 2022. https://www.nice.org.uk/guidance/ng212/chapter/Recommendations
• Scottish Government. The Scottish Health Survey 2022, volume 1: main report, chapter 1 Mental Health and Wellbeing. https://www.gov.scot/publications/scottish-health-survey-2022-volume-1-main-report/pages/5/
• Scottish Government. The Scottish Health Survey 2024, volume 1: main report, chapter 1 Mental Health and Wellbeing. Published September 2025. https://www.gov.scot/publications/scottish-health-survey-2024-volume-1-main-report/pages/1--mental-health-and-wellbeing/
• Stewart-Brown S, Tennant A, Tennant R, Platt S, Parkinson J, Weich S. Internal construct validity of the Warwick-Edinburgh Mental Well-being Scale (WEMWBS): a Rasch analysis using data from the Scottish Health Education Population Survey. Health and Quality of Life Outcomes. 2009;7:15. https://pmc.ncbi.nlm.nih.gov/articles/PMC2669062/
• Tennant R, Hiller L, Fishwick R, Platt S, Joseph S, Weich S, Parkinson J, Secker J, Stewart-Brown S. The Warwick-Edinburgh Mental Well-being Scale (WEMWBS): development and UK validation. Health and Quality of Life Outcomes. 2007;5:63. doi:10.1186/1477-7525-5-63
• University of Warwick. WEMWBS, how to use and score the scales, and WEMWBS licences and pricing. https://warwick.ac.uk/services/innovations/wemwbs/how/ and https://warwick.ac.uk/services/innovations/wemwbs/licenses/ (accessed 17 October 2026)