Temporal Trends and Service Variation in the Relative Cost of Dying: Canada 1991-2001

Abstract

Objective: To measure time trends in health care costs for individuals over 65 years of age based on individuals' survivor status for four different categories of health care services.

Data: Patient-level utilization data for four categories of publicly insured health services obtained from the British Columbia Linked Health Database. Data include hospital services, continuing care (including nursing homes and home care), prescription medications, and physician billings. The data are comprehensive, comprising over 99% of the provincial senior population over the period 1991 to 2001.

Study Design: This was a descriptive study where costs were either directly obtained from the data source or calculated by applying unit service costs to recorded utilization. Inflation-adjusted costs were tracked over the study period for decedents (dying in a given year) and survivors of the same age. The ratio of costs between decedents and survivors was measured for different age groups and different health service categories. The share of spending attributed to decedents was measured, as was the share of decedent and survivor spending in each service category.

Principal Findings: Decedent costs rose by 10% or less over the study period, falling initially in a policy environment of spending cuts, and then rising again as budgets were increased. In contrast, survivor costs fell significantly, particularly for hospital and continuing care costs. The youngest age group experienced the largest drop in survivor costs for continuing care (down 32%), while the largest drop for survivor hospital costs occurred in the oldest age group (down 29%). Both decedent and survivor costs for medical billings were unchanged, but the cost of prescription drugs rose significantly (almost 50%). For all services combined, survivor costs fell between 3% (ages 66-70) and 14% (ages 91+). As a result of the diverging trends in decedent and survivor costs, the decedent/survivor ratio grew by more than 10%over the study period. The highest growth in the ratio took place among individuals aged 81-90, expanding from 3.1 times in 1991 to 3.8 times in 2001. Although the overall death rate among seniors in BC fell slightly during the study period, the proportion of costs allocated to decedents rose. The results differ from previous observations concerning the cost of dying, which have generally found the relative cost of dying to be stable or falling slightly, and a declining share of costs attributable to decedents.

Conclusions: The increase in the cost of dying is likely due to a combination of policy effects and changes in population health. Falling survivor costs coincide with evidence that morbidity rates among the elderly have improved at a faster rate than mortality, so the average health care requirement among survivor populations may be declining. Furthermore, growth in expenditure categories oriented towards prevention and maintenance (such as prescription drugs and doctor visits) may be paying off in lower utilization of more expensive remedial care categories such as hospitals and continuing care. The data give some cause for concern that rationing of services is taking place in continuing care, particularly among the most elderly cohorts.

Keywords: Cost of dying, health expenditures, aging, demographics