An investigation into the psychometric performance of existing measures of health, quality of life and wellbeing in older adults

The UK population is ageing, with the proportion of the population aged over 65 continuing to rise. Old age is associated with increasing prevalence of frailty, characterised by a slow and steady decline in health and functioning. Older adults experiencing frailty often have complex needs, requiring complicated combinations of health and social care services over the long-term. This has led to a push for integration between health and social care services. Traditionally, the economic evaluation of health services has focussed on health-related quality of life (QoL) as the unit of benefit. However, current health measures may not be appropriate to evaluate the outcomes of social care or integrated health and social care interventions. Outcome measurement methods may need to be adapted to include broader assessment of QoL or wellbeing. This thesis aims to examine the psychometric performance of existing measures of health, QoL and wellbeing in older adults. First, the existing evidence of the psychometric performance of the EQ-5D-5L, SF-12v2, ASCOT, WEMWBS and ONS-4 was systematically reviewed. Then, item response theory (IRT) was used to examine the construct and structural validity and internal consistency of these measures in older adults aged 65+. Differential item functioning (DiF) analyses assessed whether older and younger adults with the same underlying health, QoL or wellbeing had different expected scores, indicating bias due to age. Lastly, the content validity of the EQ-5D-5L, SF-12v2, WEMWBS and ONS-4 was explored using cognitive interviews in older adults. This thesis identified some key findings which can inform the choice of measure in the evaluation of health and social care interventions aimed at older adults. In the IRT study, the ASCOT and EQ-5D-5L displayed ceiling effects, while the SF-12v2 and EQ-5D-5L exhibited DiF, both of which can bias the estimates of effectiveness obtained from these measures in an economic evaluation. The cognitive interviewing study provided insight into the way older adults conceptualise QoL and how this impacts the way they respond to items. Issues with response shift were broadly identified, which are the likely cause of DiF. Participants found the functional focused EQ-5D-5L items easier to answer and mostly relevant to their situation. The relevance of broader subjective wellbeing items on the WEMWBS and ONS-4 and negatively phrased mental health items across the measures were commonly questioned as these concepts were not prioritised in older adults’ conceptualisation of QoL. However, the coverage of any of the measures would need to be extended to include broader elements of QoL identified as important to older adults. This may be through adaptations to the EQ-5D-5L, such as bolt ons, or the development of a new measure, possibly based on the style of the EQ-5D-5L.