Self-rated versus Caregiver-rated Health for Patients with Mild Dementia as Predictors of Patient Mortality

Objective: Self-assessment of health is a strong and independent predictor of mortality for cognitively intact people. Because the ability of patients with dementia to rate their own health is questionable, caregiver-rated health for patients may serve as a proxy. The authors aimed to validate and compare self-and caregiver-rated health for patients with dementia as independent predictors of patient mortality. Methods: This was a post-hoc analysis of data from The Danish Alzheimer's Disease Intervention Study, a randomized controlled trial of psychosocial intervention for 330 patients with mild dementia and their caregivers with a 36-month follow-up. Patients and caregivers rated patients' health on the Euro Quality of Life Visual Analog Scale (EQ-VAS) from 0 (worst) to 100 (best). The ability of self-and caregiver-rated health for the patient to predict patient mortality was analyzed as hazard ratios (HRs) from Cox proportional hazard regression models, controlling for age, depression, comorbidities, functional level, quality of life, and randomization group. Results: Compared with the highest scores of 80-100, caregiver-rated EQ-VAS scores < 50 had an HR of 9.8 (95% CI: 2.9-33.1), scores between 50 and 64 had an HR of 3.8 (95% CI: 1.2-12.3), and scores between 65 and 79 had an HR of 4.6 (95% CI: 1.4-14.7). Self-rated EQ-VAS did not statistically significantly predict mortality. Conclusion: Caregiver-rated health for patients with mild dementia using the EQ-VAS was shown to be an independent predictor of patient mortality with a dose-response pattern but patient-rated EQ-VAS was not.