OBJECTIVES: Decision makers increasingly recognize that innovative health interventions not only benefit patients but also have spillover effects on caregivers. However, there can be logistic challenges to obtaining assessments from both patients and caregivers. This study investigated suitability of patients as proxies by examining agreement between caregiver self and proxy assessments of caregiver HR-QoL (by patients) using the EQ-5D-5L.

METHODS: Data was collected from 504 caregiver-patient dyads between August 2022 and February 2023 using a US-based cross-sectional online Qualtrics panel survey. Caregiver HR-QoL was assessed by caregivers and patients using the EQ-5D-5L and EQ-VAS. Directional bias was assessed using mean difference scores and strength of agreement was quantified 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).

RESULTS: The mean (SD) age of caregivers was 49.2 (±15.4) and that of patients was 62.7 (±18.9) years. Dyads were most commonly spouses/partners (34.5%). Fair to moderate agreement between caregivers’ self and patients’ proxy assessments was observed across five dimensions in EQ-5D-5L (ICC ranging from 0.37 [Usual Activities] to 0.55 [Pain/Discomfort]), index-based score (ICC=0.54) and VAS score (ICC=0.57), whereas moderate to excellent agreement was observed between patients’ self and caregivers’ proxy assessments (ICC = 0.59 to 0.82). Notably, patients tended to overestimate caregivers’ HR-QoL. These results did not differ based on relationship, time of caregiving, and baseline HR-QoL.

CONCLUSIONS: Results from this study suggested patients as proxies for caregivers are slightly less reliable than caregivers serving as proxies for patients but may still be sufficient to serve as proxies. The threshold for acceptability where patients can serve as proxies for caregivers or at least provide valuable insight into spillover effects in terms of health outcomes using the EQ-5D and other measures of health warrants further research as a mechanism for capturing broader health effects of novel health technologies.