Differences in Proxy Versus Self-Report: Assessing Health-Related Quality of Life and Functional Status in Older Adults

This study examines differences in health-related quality of life and functional limitations for older adults (ages >65) who self-report versus those with proxy respondents.  Past research has shown that older adults who use proxy respondents have a higher incidence of cognitive impairment and lower functional status than their peers.  They tend to rate their quality of life and functional status higher than proxies.  Using data from the Medicare Health Outcomes Survey, we found support for these trends. Unadjusted, self-reporting older adults report on average 4.94 physically unhealthy days (PUD) and 2.38 mentally unhealthy days (PUD) in the previous 30 days. These numbers are significantly lower than the number of days reported by older adults who had a family member (mean PUD = 11.07, mean MUD = 7.55), friend (mean PUD = 8.81, mean MUD = 5.83), or professional caregiver (mean PUD = 10.99, mean MUD = 8.06) complete the survey on their behalf.  However, it is difficult to disentangle whether these effects are based on severity of physical or cognitive impairment, or perceptual difference.  It is very likely that individuals that used a proxy reporter are in worse health, but there are also reasons to believe that older adults perceive health conditions impact on their quality of life differently than their friends and family members. In order to better answer these questions, we employed propensity score matching techniques to pair older adults who completed the Medicare Health Outcomes Survey using proxy reporters (n = 7,961) with older adults who completed the survey themselves (matching pool, n = 75,270). Adults were matched based on demographic factors (e.g., age, gender, education, race, marital status) and being diagnosed with any of 13 different chronic conditions (e.g., hypertension, heart attack, stroke, arthritis, sciatica, diabetes, cancer, emphysema).  The matching procedure resulted in pairs with very similar demographic and chronic condition profiles, yet the gap between self-report and proxy-report persisted.  Moreover, while gaps in self-reported HRQOL were reduced after matching, self-reporting adults reported fewer PUD (6.19) and MUD (3.14) compared to the proxy-report group (PUD=10.82; MUD=7.29). Additionally, significant and clinically important differences persisted in reporting of functional status.  For example, even after matching, only 14% of the self-reporting sample stated that they have difficulty dressing, compared to 38% of the proxy reporters, and self-reporters.  We address the various clinical, research-oriented, and community-level intervention implications of these findings, which revolve around improving quality of life for older adults who are unable to self-report due to cognitive, physical, or communicative barriers.