Comparing patient- and proxy-reported health-related quality of life of frail older patients at the emergency department

Abstract

PurposeTo examine agreement between patient- and proxy-reported Health-Related Quality of Life (HRQoL) scores, and their associations with clinical measures and patient demographics.MethodsWe examined the patient-proxy agreement in HRQoL within a cohort of community-dwelling frail older patients (Clinical Frailty Scale [CFS] score 4–6) planned for discharge from the Emergency Department (ED) of a 1700-bed tertiary hospital in Singapore. Patients were recruited as part of the Emergency Department Interventions for Frailty (EDIFY) program, a quasi-experimental study evaluating the effectiveness of a multicomponent frailty intervention (MFI) in maintaining or improving functional outcomes. HRQoL was measured using the EuroQoL five-dimensional questionnaire (EQ-5D-5L) for patients and proxies. We analysed a sample of dyadic-reported HRQoL to explore the inter-rater agreement using intraclass correlation coefficients (ICC), and the influencing factors of concordance.ResultsWe recruited 140 participants with 23 pairs of patient-proxy dyads having concurrent EQ-5D reported. There was poor inter-rater agreement between patient- and proxy-reported EQ-5D (ICC 0.18) with proxies perceiving poorer average HRQoL than patients. On the domain level, there was poor agreement in self-care and pain domains (ICC − 0.07 and 0.20), where proxies perceived more problems than patients. In multivariate regressions, lower frailty scores were associated with poorer agreement in EQ-5D scores (p < 0.05).ConclusionIn conclusion, our findings demonstrate differences between patient- and proxy-reported HRQoL outcomes among frail older patients, with significant discordance observed among patients who are “less frail”. Hence, proxy-reported responses should be interpreted with care especially when the abovementioned characteristics are present in an older patient.