Mental well-being: An important outcome for mental health services?

Summary

Mental well-being is being used as an outcome measure in mental health services. The recent Chief Medical Officer's (CMO's) report raised questions about mental well-being in people with mental illness, including how to measure it. We discuss whether mental well-being has prognostic significance or other utility in this context.

The World Health Organization defines mental well-being as an individual's ability to develop their potential, work productively and creatively, build strong and positive relationships with others and contribute to their community. ¹ This view distinguishes subjective happiness or life satisfaction (hedonic well-being) from positive psychological functioning (eudaimonic well-being). The mental well-being literature can be confusing as many similar-sounding terms are used interchangeably: social or mental capital, positive mental health, psychological or subjective well-being. The WHO definition of mental well-being is concerned exclusively with positive mental health states, and this approach is also evident in the way that terminology is used in UK policy documents. Nevertheless, it is sometimes unclear whether the term ‘mental well-being’ implies the absence of mental illness or distress. Well-being has been trumpeted as a measure of national prosperity, and linked to improved physical and mental health. It has been identified as a public health target and criterion for commissioning and assessing mental health services. ² But questions remain about the relationship between mental illness and mental well-being, and about the potential for diverting resources away from evidence-based treatments for mental disorders. These issues were highlighted in the recent Chief Medical Officer (CMO) report on public mental health that challenged the empirical grounds for extending mental well-being into clinical commissioning and argued against mental well-being ‘receiving priority funding over better established fields, including quality of life’. ²

Mental well-being and mental distress

Mental disorders are characterised by psychopathology, distress and impaired functioning. Huppert ³ and others argued that mental disorders (‘languishing’) and mental well-being (‘flourishing’) were opposite ends of a single dimension. However, further work has shown that, although correlated, mental illness and mental well-being are independent phenomena. Secondary analysis of data on over 7000 adults from the 2007 Adult Psychiatric Morbidity Survey (APMS) demonstrated that associations with well-being scores were not significantly altered by adjusting for comorbid mental disorder. ⁴ These findings were consistent with those from other studies that indicate that mental well-being is more than just the absence of mental illness symptoms and distress, and that (although correlated) mental well-being and mental distress are independent of one another. The APMS findings also showed that at least moderately high levels of well-being may be achieved in the context of mental illness, which is salient when considering whether mental well-being should be a routine outcome measure in mental health services. ⁴ Evidence detailed later in this editorial also supports this conclusion. However, we know less about the determinants and variability of mental well-being among those who experience mental health problems than in the general population. As mental illnesses typically relapse and remit, mental well-being may vary with the phase of illness and the number, frequency or duration of relapses.

Measuring mental well-being in people with mental illness

Evaluating interventions to improve mental well-being in people with mental illness depends on valid measurement, but there is only limited evidence to guide the assessment of mental well-being in this context. ² This is a significant barrier to studying mental well-being and its potential determinants in people with mental illness. ² Since mental well-being is a state of positive mental health, measures should comprise positively phrased items, such as those which make up the Warwick-Edinburgh Mental Wellbeing Scale (WEMWBS), ⁵ WHO-Five Well-Being Index (WHO-5) ⁶ and the Satisfaction with Life Scale. ⁷

Although generic measures of mental well-being have been used for people with mental illness, their validity in these populations has rarely been evaluated; we do not know whether responses to generic mental well-being items may be biased by the experience of past or current mental illness. Only the WHO-5 has been validated in English in mental illness, specifically in affective and anxiety disorders. ⁸ The Subjective Well-being under Neuroleptic Treatment Scale (SWN) ⁹ was developed for people with schizophrenia receiving antipsychotics. However, one-half of this scale comprises negatively worded items and it covers domains that are not central to mental well-being, including physical functioning. WEMWBS, despite being recommended by healthcare organisations for measuring mental well-being in the context of mental illness, has only been validated in non-clinical populations in the UK.

Mental well-being and mental health services

The 2011 UK government document No Health without Mental Health emphasised mental well-being as an important service outcome as part of patient-centred, recovery-focused care. ¹⁰ However, judging services according to mental well-being outcomes rather than changes in symptoms and disability is not self-evidently consistent with their traditional mission: the consequences of doing so need to be considered carefully. Measuring mental well-being routinely may alter therapeutic relationships in unintended ways. There is a risk that in prioritising mental well-being, professionals might be excused from achieving more challenging outcomes, namely alleviating symptoms and reducing disability. ²

We would argue that two conditions must be met to justify the routine assessment of mental well-being among mental health service users. First, evidence is needed that mental well-being modifies the risk of onset, recovery from or recurrence of episodes of mental illness; in other words that it has prognostic significance in terms of mental health, social functioning or use of healthcare. Second, it must be shown that mental well-being is independent of mental illness and social functioning and therefore unlikely to be captured by measures that assess either of these phenomena.

Mental well-being and mental illness

Although the behavioural and psychosocial determinants of mental well-being may not necessarily resemble those of mental illness, mental well-being is associated with specific forms of psychopathology – examples are discussed below. However, the evidence base is generally limited by substantial methodological variation (including the use of different and often unvalidated measures of mental well-being) and a dearth of longitudinal studies, inhibiting understanding of cause and effect. ²

Anxiety and depression

Maintaining high levels of mental well-being is likely to be difficult in the presence of symptoms of anxiety and depression. However, recent longitudinal data demonstrate that this may be more complicated than (simply) covariance. A recent study of over 1000 Australian in- and day patients with depression or anxiety demonstrated that an intervention (giving feedback during psychological treatment) improved depressive symptoms but not mental well-being, ¹¹ supporting the view that these are independent outcomes.

Insomnia

There is a wealth of cross-sectional evidence linking sleep problems and mental well-being, but less robust evidence of longitudinal associations. A small, prospective study of 75 university students ¹² found no significant prospective improvements in life satisfaction among those whose sleep increased in duration or quality over 3-month follow-up. Those who reported a reduction in daily sleep quality over 3 months were significantly more likely to report a reduction in life satisfaction (P<0.01). ¹² Nonetheless, poor mental well-being in the context of sleep problems may not be associated with greater need for psychiatric care. A cross-sectional general population study of over 8000 Australians found that although the 5% with insomnia were significantly more likely to have poor mental well-being (odds ratio (OR) = 2.34, 95% CI 1.11–4.93) and visited their general practitioner more often (OR = 1.57, 95% CI 1.06–2.33), insomnia was not significantly associated with use of psychotropic medication or hospital admission. ¹³

Delusions and hallucinations

Mental well-being is inversely associated with psychotic symptoms. In 83 out-patients with schizophrenia, psychotic symptoms were negatively correlated with quality of life, but interestingly this association was confounded by insight, ¹⁴ demonstrating the complexity of the relationship between mental well-being and mental illness. Among people with first-episode psychosis, admission to hospital was associated with better quality of life ¹⁵ suggesting that illness severity per se may not automatically predict well-being; better mental well-being might also reflect the quality and intensity of care received.

Social functioning and healthcare use

Social functioning is correlated with psychopathology but may be independent of mental well-being. Psychiatric out-patients with serious mental illness in remission demonstrated higher functioning scores but not higher well-being compared with similar patients not in remission, although this used the limited SWN to measure mental well-being. ¹⁶

Healthcare use and mental well-being may also be independent. A 2-year structured rehabilitation programme for those with serious mental illness led to improved quality of life and psychosocial functioning in those who met their rehabilitation goals v. those who had not. However, there were no significant differences in healthcare use between the two groups at 2-year follow-up. ¹⁷

Should mental well-being be used to support the commissioning and delivery of mental health services?

Valid methods of evaluating healthcare interventions are required to support payment by results, and National Health Service providers are required to collect patient-reported outcomes and experiences in part to prevent ‘gaming’ to maximise income. Mental well-being could serve as a patient-rated outcome measure, but the dearth of validated measures in people with serious mental illness remains a major concern. The CMO has sensibly encouraged policy makers and commissioners to heed the uncertainty surrounding mental well-being, warning that ‘wellbeing policy is running ahead of the evidence’. ² However, existing evidence suggests that symptomatic and functional outcomes, needs for care and service use appear to be independent of mental well-being to varying degrees. Therefore, mental well-being is not captured completely by existing measures of these states. Mental well-being also has strong conceptual resonances with recovery from mental illness, including notions of hope, purpose and fulfilment, and may be similarly valued by patients. Taken together, these could represent significant arguments for mental well-being as a distinct service outcome in its own right. However, the utility of measuring mental well-being routinely in mental health services has not yet been established. Further research is needed to validate measures of mental well-being in people with serious mental illness, determine the usefulness (and costs) of routinely measuring mental well-being in this population, and to explore the views of patients on the relative importance attached to different service outcomes.

Conclusions

The place of mental well-being in the delivery of mental healthcare remains uncertain and the CMO has stated categorically that this should not be part of current clinical commissioning. Nevertheless, mental well-being is an important public health heuristic and has clear resonances with concepts underpinning recovery from mental illness. The evidence base linking mental well-being and mental illness remains poorly developed, but we believe that two conditions for measuring mental well-being in mental health services have been at least partly met. It appears that mental well-being may be associated with onset, recovery and/or recurrence of episodes of mental illness although the actual detail of these associations is complex; and that it is at least partly independent of symptoms, social functioning or need for mental healthcare. Mental well-being is not fully captured by measures of these phenomena.

However, there are two important caveats. First, it is essential to validate measures of mental well-being in people with serious mental illness, and to know more about the (relative) value that patients place on mental well-being as a service outcome. And second, mental well-being must not be allowed to supersede other outcomes and obscure the imperative to deliver the most effective evidence-based treatments to those with mental illness.