Socioeconomic Status, Health Status, and Healthcare Costs: Can Medical Savings Accounts Meet the Healthcare Needs of Canadians?

Key Implications for Decision Makers

The purpose of this project was to determine whether medical savings accounts could meet the healthcare needs of Canadians while making people financially accountable for their decisions about the healthcare services they use. The research team wanted to know whether it would be possible for an individual to save in low-cost years in order to pay for additional needs in high-cost years, particularly when it comes to implementing a medical savings account for prescription drug purchases.

  • Individuals of the same age and sex use vastly different amounts of hospital care, physician services, and prescription drugs. Even if we control for health status, it is very difficult to determine the needs of an individual based on objective criteria. In any given year, many people spend far more than expected and others far less.
  • Individuals with high costs in one year tend to have high costs in later years. This means they would not have enough low-cost years when they could save for future needs.
  • Individuals control their use of prescription drugs and their visits to doctors more than their use of hospital services, where decisions are based on clinical criteria. However, costs of physician services and prescription drugs vary least over time. Therefore high users, most of whom are sick, will be penalized year after year, while low users, most of whom are healthy, will build up large surpluses over time.
  • Costs for prescription drugs are particularly steady over time. High users remain high users, and vice versa. Therefore, medical savings accounts may be most problematic for this area.
  • People with low incomes are more likely to incur high hospital costs at every age, reflecting their poorer health status.
  • The relationship between physician and prescription drugs costs and income is more complicated. People with low incomes have poorer health than higher-income people. Nevertheless, they do not use as many physician services or prescription drugs as one might expect, taking into account their poorer health.

Executive Summary

Many people argue that Canadians overuse healthcare resources because they don't pay for physician services or hospital care. Some have suggested that the best way to encourage Canadians to make responsible choices is to introduce medical savings accounts. Individuals would receive an allowance based on their needs and would use this account to pay for the services they use. They would save money in years when their needs were lower than average in order to pay higher costs in years when their needs are greater.

We wanted to know if we could use objective criteria (such as age, sex, and income) to determine how much a person would need to spend on healthcare in an average year. We suspected that, in any given year, some people would spend far more and others far less than average for their age and sex, even if we attempt to control for health status. We wanted to know whether it would be possible for an individual to save in low-cost years in order to pay for additional needs in high-cost years. If some people consistently have high needs, this would not be possible.

We used the Manitoba sample of the 1996 National Population Health Survey. Interview data from individuals in the sample who gave permission were linked to health-use records and vital statistics. We accessed complete out-of-hospital prescription drug records, records of all contacts with physicians, summaries of all hospital care, household income data, and date and cause of death for 8,748 individuals.

We discovered that people of the same age and sex use vastly different amounts of hospital care, physician care, and prescription drugs; various ways of controlling for health status do not remove the difference. This means that if medical savings accounts were implemented, some people would spend far less than their allowance and others far more in any given year.

We also found that spending, particularly on physician services and prescriptions, are strongly related over time. This means that individuals who spend more than expected this year will probably spend more than expected in subsequent years. Individuals cannot save in low-cost years to cover their needs in high-cost years, as medical savings accounts assume, because the healthcare costs of an individual do not balance out over time.

Medical savings accounts aim to make individuals personally accountable for their healthcare use. Individuals have greatest control over their use of physician services, because in most cases they make the first contact with their family doctor. Individuals also have control over their use of prescription drugs: they make decisions about whether to accept recommendations from their physician, and they are responsible for filling prescriptions and following regimens. They may also suggest drugs to physicians, who may or may not prescribe them. Hospital admissions, on the other hand, are largely out of the control of individuals. Medical savings accounts are most problematic in the case of prescription drugs and physician services, because these costs tend to be steady over time.

It is well-documented that people with low incomes have poorer health than people with higher incomes. Because hospital and physician care is provided at no cost to individuals, and prescription drugs are largely subsidized for low-income individuals, we expected to find low-income people incurring consistently higher costs.

While we found higher hospital costs for low-income people, reflecting their poorer health status, we did not find a direct relationship between physician costs and income, or between prescription drug costs and income. There are two possible explanations: either low-income people use too few prescription drugs and too little physician care, possibly resulting in higher hospital costs later as chronic conditions are not well-managed, or high-income people use too many prescription drugs and physician care, given their health status. Clearly, the appropriate policy response depends very significantly on how and why income affects hospital, physician, and prescription drug costs.