Value-based care. The enormous number of conference presentations, boardroom discussions, policy documents and research publications on this topic suggests that many health system stakeholders see it as an important shared objective. If we truly are on the road to value-based reimbursement clarity is needed on how value is defined and measured.
Value = outcome/cost is a conceptually simple formula. But we should ask: which outcomes, and from whose perspective?
In a patient-centred health care system, the outcomes we care about the most are (or should be) the outcomes that matter most to patients, including the ones only patients can report: symptoms, function, and health-related quality of life (HRQoL). Patient-reported outcome measures (PROMs), therefore, are key metrics of health care system performance, whether used for ‘judgement’ (e.g., to set payment) or for learning and improvement. PROMs increasingly find a role in the clinic too.
Clinical outcomes, like hospital-acquired infections, surgical complications, medication errors and stroke or heart attack rates, are, naturally, still highly relevant and need to be measured, along with costs, and PROMs. We would expect clinical outcomes to correlate with patient-reported ones.
What is health-related quality of life, and how is it measured?
EQ-5D is probably the most used HR-QoL metric. It assesses quality of life in five dimensions – mobility, self-care, usual activities, pain/discomfort, and anxiety/depression. Preference weights convert EQ-5D into a single, multi-dimensional value that can then be used to assess health economic outcomes, e.g., for Health Technology Assessment (HTA).
HTA can be considered a special case of value-based care where value is assigned objectively, for example, to a new drug, imaging procedure or surgical technique used in a specific context. EQ-5D has been adopted by HTA agencies such as NICE in the UK and ZIN in the Netherlands and appears in many HTA guidelines. It has good psychometric properties and is available in numerous approved translations, including Afrikaans, isiXhosa, and isiZulu. EQ-5D is free for research purposes, but other uses have a license cost.
Does an HR-QoL measure adequately represent all aspects of health that an individual would value? There are 3,125 (55) health states in the 5-level version of EQ-5D, yet a criticism of EQ-5D is that it fails to capture some important elements of health and benefits of interventions.
In the social and personal care sectors, for example, autonomy, social participation, and meaning are important. For social care, palliative care, and the treatment of many chronic conditions (e.g., diabetes, dementia, and spinal cord injury), improved physical or mental health are not the only desired outcome. Dignity, autonomy and control, satisfaction with relationships, and the ability to participate in meaningful activities are valued.
While building a Health Risk Assessment (HRA) tool for a UK occupational health care company recently, we had to carefully consider what health really means. The World Health Organisation definition of health includes the concept of well-being, not just the absence of disease. Authorities in the UK and elsewhere know that social determinants drive health as much as medical care does. As the population ages, the number of years people spend in a state of frailty and ill-health is increasing, so the importance of social care grows. New models of care are being introduced to facilitate integration across acute, primary, mental health, specialist, and social care services, and the effects of these interventions need to be measured. Interest is therefore rising in quantifying concepts like well-being and capability, which capture benefits relevant to health economic decision-making.
The HRA tool aims to assess physical and mental health status, as well as various personally actionable health-related factors like diet, exercise, smoking, alcohol consumption, and the use of available preventive care like screening tests and vaccination. In assessing health status, and the risks contributing to poor health, we assume clients – and society in general – are interested in what a person can do or be, i.e., capabilities valued for their own sake – their contribution to quality of life. Complexity arises from the fact that these same capabilities can help to drive physical and mental health.
Capability, opportunity and motivation (“COM-B”) are therefore important measurable factors that are not health outcomes per se but three important drivers of behaviour affecting things like eating better, exercising more, and stopping smoking, that we expect to lead to better physical health and therefore include in the HRA.
We choose, in addition, to measure aspects of subjective well-being (SWB), the tools for which include multi-item measures of psychological well-being like the Warwick-Edinburgh Mental Well-Being Scale. These provide an overall indicator of quality of life, which could be used in health care and other sectors.
The ICECAP-A measure attempts to form a generic list of quality-of-life attributes suitable for economic evaluation. Five items define states that have been valued using a preference elicitation technique, namely:
- feeling settled and secure.
- love, friendship, and support.
- being independent.
- achievement and progress.
- enjoyment and pleasure.
Another measure that includes well-being alongside health dimensions is the Assessment of Quality of Life (AQoL) instrument, developed in Australia and used by Voice of the Patient there, which covers five psychosocial aspects: mental health, happiness, self-worth, relationships, and coping.
 Michie, et al. (2011). The behaviour change wheel: A new method for characterising and designing behaviour change interventions. Implementation Sci, vol. 6 (1). doi:10.1186/1748-5908-6-42.
New measures like ICECAP-A and AQoL take longer to complete than the EQ-5D. Agencies seeking to compare or value outcomes can’t only use a health care sector-specific measure of quality of life like EQ-5D; adding other items makes things more complicated.
Patient-reported outcome measures provide a way to measure what matters to people. Finding ways to include them in economic value metrics and payment mechanisms is a challenge worth solving. Appropriate measurement should facilitate positive change in health care systems and related social systems that impact health and quality of life.
For more information visit:
Brazier JE, et al. Future Directions in Valuing Benefits for Estimating QALYs: Is Time Up for the EQ-5D? Value in Health 22 (2019) 62-68.