Elder‐Proxy Agreement Concerning the Functional Status and Medical History of the Older Person: The Impact of Caregiver Burden and Depressive Symptomatology

OBJECTIVES: To examine the influence of caregiver burden and depressive symptomatology on elder-proxy response concordance regarding the older person's functional status and medical history. DESIGN: Cross-sectional study via telephone interviews. SETTING: Community-dwelling older people and caregivers in North Carolina. PARTICIPANTS: 340 matched pairs of frail persons aged 65 and older and their respective caregivers. MEASUREMENTS: Multidimensional Functional Assessment: The OARS methodology RESULTS: Percent agreement on the ADL items ranged from 97.6% on personal hygiene to 99.7% for toileting, with moderate kappa coefficients. IADL percent agreement ranged from 71.5 to 93.7%, with fair to moderate kappa coefficients. Agreement among the medical history items ranged from 76.3 to 98.5% (kappa = .138-.831). Response bias for the IADL composite measure is influenced marginally by caregiver burden (F259 = 1.751, P = .098). Five of the individual IADL bias items are influenced significantly by burden, such that an increase in burden results in a greater likelihood that the caregiver will overstate disability compared with the rating by the older person. Response bias on the ADL scale was increased among persons who experienced more caregiver burden (OR = 1.096, 95% CI = 1.000, 1.192) and those who spent more hours providing care (OR = 1.012, 95% CI = 1.001, 1.024). Additionally, black caregivers were more likely than white caregivers to disagree with the older people on the ADL scale (OR = 2.73, 95% CI = 1.642, 3.809). A composite of the medical history items is influenced by the relationship of the caregiver to the older person; bias is more likely among adult children ((F227 = 1.56, P = .081). CONCLUSION: Elder-proxy concordance is highest among ADL items, followed by medical history items and IADL items. Caregiver depressive symptomotology had no significant impact on elder-proxy response concordance on any of the three outcomes of interest: IADL and medical history bias and ADL disagreement. However, caregiver burden was marginally predictive of bias on the total ADL and IADL scales. Additionally, increased burden was significantly predictive of bias on five of the seven individual items of the IADL scale, suggesting that the more burden a caregiver feels, the greater likelihood that s/he will overstate the older person's disability compared with self-report. These findings suggest that clinicians and researchers who use proxy reports to determine treatment regimens and complete data collection efforts may do so with confidence on ADL individual items and medical history items when the older person's frailty is marginal. However, caregiver burden may result in misleading representation of the older person's functional status, specifically in regard to IADL items.