To evaluate the suitability of care recipients as proxies for assessing caregivers’ health-related quality of life (HRQL) using the EuroQol 5-Dimension 5-Level instrument (EQ-5D-5L).
A cross-sectional online survey of 504 US caregiver-patient dyads was conducted. Self and proxy versions of EQ-5D-5L and EuroQol Visual Analog Scale (EQ-VAS) were completed by both caregivers and care recipients. Agreement was evaluated using intraclass correlation coefficients (ICC), interpreted as poor (0-0.2), fair (0.2-0.4), moderate (0.4-0.6), substantial (0.6-0.8), and excellent (>0.8). Directional bias was assessed through mean differences between proxy and self-assessment score.
Care-recipient proxy assessments showed fair-to-moderate agreement with caregiver self-assessments across EQ-5D-5L dimensions (ICC range: 0.37-0.55) and summary scores (EQ-5D-5L Index: ICC = 0.54, EQ-VAS: ICC = 0.57). Care recipients consistently overestimated caregivers’ HRQL across all dimensions, with the largest differences in usual activities and anxiety/depression (both mean difference = 0.24). The directional bias for the EQ-5D-5L Index showed systematic overestimation (mean difference = −0.06), with a small effect size (0.22). In comparison, caregivers showed higher agreement when assessing care recipients (ICC range: 0.59-0.82) with slight underestimation (mean difference = 0.03, effect size = 0.09). These patterns remained consistent across subgroup analyses.
Care recipients showed overestimation of caregiver HRQL, although the effect size was small. Although the agreement was lower than in traditional caregiver proxy assessments, the small magnitude of bias suggests that this approach could be valuable for group-level assessments. This method may offer a practical solution for capturing caregiver spillover effects in health technology assessments when direct caregiver assessment is challenging.