Table 1 Discriminant construct validity hypotheses, by QoL measure

ASCOT-Proxy-Resident

Functional dependence

No significant (or significant low negative) association based on previous research with care home residents in England [8, 16, 18, 28].

Y

Cognitive impairment

Significant negative association based on previous research with care home residents in England [8, 16, 18, 28].

Y

Potentially avoidable emergency hospital admission

Significant negative association based on measurement construct, which is designed to capture quality and effectiveness of care [6, 46].

Y

Care home quality

Significant positive association based on previous research with care home residents in England [14–15] and measurement construct (i.e. SCRQoL), which is designed to capture the quality and effectiveness of care [6, 46].

Y

ICECAP-O

Functional dependence

Significant negative association based on previous research with care home residents internationally [19,20,21].

Y

Cognitive impairment

Significant negative association based on previous research with care home residents internationally [19,20,21].

Y

Potentially avoidable emergency hospital admission

No significant association based on measurement construct, which is designed to capture capability wellbeing rather than quality and effectiveness of care [9–10, 47].

Y

Care home quality

No significant association based on measurement construct, which is designed to capture capability wellbeing rather than quality and effectiveness of care [9–10, 47].

Y

EQ-5D-5L Proxy 2

Functional dependence

Significant negative association based on previous research with care home residents internationally [16, 18] and measure construct, which includes a rating of ‘usual activities’. This overlaps with the construct of functional independence, as measured by the Barthel Index [48–49, 52].

Y

Cognitive impairment

No significant association based on previous research with care home residents internationally [26–27].

N

Potentially avoidable emergency hospital admission

No significant association based on instrument construct, which is not designed to capture care quality and effectiveness [48–49].

Y

Care home quality

No significant association based on instrument construct, which is not designed to capture care quality and effectiveness [48–49].

Y

All QoL measures

Functional dependence

EQ-5D-5L Proxy 2 will have a stronger sized negative association than ICECAP-O, and ICECAP-O a stronger sized negative association than ASCOT-Proxy-Resident, based on measure constructs [6, 9–10, 46,47,48,49].

Y