The purpose of this assignment was to use data on health care spending and outcomes in countries throughout the world as a basis for interactive tables and charts. There were four basic questions to be investigated:
Obviously, these questions will be answered in different ways, depending on how things like spending on health care, and health care outcomes, are defined. I describe the sources of data and the definitions I used, below.
We were provided with a data set that included eight different measures of spending on health care from 193 countries. These data came from the World Health Organization’s Global Health Observatory Data Repository, and were all from the year 2010. In addition, the dataset included four measures of health care outcomes for each country: the total and per-capita numbers of doctors and nurses in the country, also in 2010. While numbers of doctors and nurses are one measure of the quality of the health care system in a country, I chose to download additional data from the WHO website that reflected actual health outcomes of the population. These included data on life expectancy at birth, infant mortality, immunization rates for common childhood diseases, and mortality rates from non-communicable diseases. The latter statistic reflects a number of so-called “lifestyle” diseases–such as diabetes, heart disease, and stroke–that are generally more common in wealthier countries and present the predominant health-care challenges in those countries. These data were also for the year 2010.
Specifically, I used the following four measures of health care spending:
As measures of outcomes, I used the following four variables:
Because countries differ in both their populations and the sizes of their economies, raw spending on health care is not a useful measure. And, since “countries” include both governments and the private sector, distinguishing between those two sources of spending is important. As can be seen from the intensity map and the table below, The top-spending countries vary depending on which measure is used.
There are a large number of possible health care outcomes that could be examined to answer this question. The World Health Organization collects data on a wide range of measures of mortality, morbidity, well-being, and preventative medicine, and reports them by country, region, year, and in many cases, age-class. I chose four measures that together represent health care challenges for both less-developed and more-developed countries.
Infant mortality is one of the most commonly cited measures of a country’s standard of living and basic health care. As countries develop economically, infant mortality is one of the measures likely to respond quickly to the availaility of better medical care. As can be seen in the table above, the twenty countries with the lowest infant mortality rates are again generally among the wealthiest countries of the world. Sixteen of the top twenty are European countries, joined by Singapore, Japan, Cyprus, and South Korea. One notable absence from the top twenty is the United States, which is number 41 on this list, despite being in the top ten countries in all four measures of health care spending.
For children that survive their first year, common childhood diseases such as diptheria, tetanus, and pertussis have historically been frequent killers. Another relatively inexpensive and effective improvement in health care outcomes as countries develop is the implementation of effective vaccination programs for these and other childhood diseases. The WHO data on DTP3 vaccination coverage show that the top ten countries in this regard again come from Europe. the next ten are more geographically diverse, however, including Gulf states of Bahrain and Oman; Island nations such as the Cook Islands, Sri Lanka, the Bahamas, the Seychelles, and Mauritius; and Asian countries China and Thailand. The fact that immunization programs are relatively inexpensive means that even countries with relatively low levels of government spending on health care can have 99% of their one-year-olds covered with the DTP3 vaccine. Indeed, it is not until one gets to number 32 on this list that the coverage rate drops from 99% to 98%. The United States is well down the list in this measure, but still has a coverage of 95%.
Life expectancy is a measure of both childhood and adult survival rates. While it is heavily influenced by childhood mortality (because younger age classes generally make up a larger proportion of the population), life expectancy also begins to reflect the risks of dying at older ages. So countries with more comprehensive health care programs might be expected to have longer life expectancies. In this measure too, the wealthier countries dominate, with the top twenty including 14 European countries, plus the Asian nations Japan (at the top of the list) and Singapore, as well as Canada, Israel, and New Zealand.The United States is in the top 40 in life expectancy, with a value of 79 for men and women combined, compared to 84 for Japan.
Finally, the measure of mortality rate from non-communicable diseases (NCDs) may reflect the down side of affluence. Rates of heart disease, stroke, diabetes and other “lifestyle diseases” are often higher in more developed countries. This is in part because development means fewer people die from infectious (i.e., communicable) diseases, and from the rigors of a dangerous and/or subsistence lifestyle. But it also results from the fact that dietary changes and reduced physical activity predispose wealthier populations to obesity, hypertension, and other metabolic disorders. Many countries in the data set do not have data on NCD mortality rates. However, we can see that the lowest rates of NCD mortality are actually in the same wealthy countries that have topped the other lists of outcomes. This may reflect the fact that these countries are able to provide the medications and other treatments that can counter the maladies that their wealthy economies predispose them to.
Because both spending on health care and health outcomes are complicated, multifaceted parameters, any relationship among them will not be simple. As a first pass at examining this relationship, I plotted four response variables (infant mortality, life expectancies (men and women combined), DTP3 vaccination coverage, and NCD mortality rates) against government spending on health care per-capita. In addition, each plot indicates the population size of the country by the size of the “bubble” representing it.
When government spending on health care per-capita increases, infant mortality decreases, as might be expected. However, it is also clear that there are diminishing returns on this particular outcome. Countries with per-capita expenditures at the highest end of the range do not have infant mortality rates that are appreciably lower than those of countries with per-capita expenditures of half as much. This is not to say that those additional expenditures don’t produce additional benefits in other outcomes, but it is clear that significant reductions in infant mortality can be achieved with lower expenditures. It is also true that other countries with comparable expenditures in the middle of the range still have infant mortality rates of 30 - 40 per thousand.
As with infant mortality, life expectancy improves with increased per-capita government spending, this time in a more linear fashion. This could be due to a variety of factors, however, some of which may be unrelated to government spending specifically on health care. As economies improve and living standards increase, things like cleaner air and water, better nutrition and education, and safer working environments can also contribute significantly to longer life expectancies. These increases in the quality of life are likely to be correlated with increased spending on health care, but may not result directly from it.
Vaccination against childhood diseases such as diptheria, tetanus, and pertussis is relatively inexpensive as health care interventions go, and can have a significant benefit in reducing childhood mortality. Thus we see that even countries with relatively low levels of per-capita government spending can acheive high rates of coverage for these vaccines. At the same time, some countries with relatively high rates of spending do not have high vaccination rates. This may be due to cultural factors that lead to resistance against vaccination, or to geographic and infrastructure factors that make it difficult to reach widely dispersed rural populations with an effective vaccination program.
The change in mortality rates from non-communicable diseases with increased government spending on health care is more complicated. Clearly, the countries with the highest per-capita government spending also have the lowest rates of mortality from these diseases. But the relationship at the lower end of the range is essentially flat, with a lot of scatter in the points. Thus it appears that a number of factors other than government spending are affecting non-communicable disease rates. This is not too surprising. Countries are likely to prioritize health care spending on the most critical issues first, such as infant mortality and childhood and other infectious diseases. Non-communicable diseases such as heart disease, stroke, and diabetes, are more expensive to treat, so a country needs to get beyond a certain threshold of economic development before it can devote significant resources to them.
The answer to this question hinges on one’s definition of “useful.” It is clear that increased government spending on health care is associated with significant improvements in some health care outcomes, such as infant mortality rates and immunization rates. Using these measures as outcomes suggests that there are decreasing marginal benefits to increased health care spending, and countries could have equally high success with less spending. On the other hand, other health outcomes, such as mortality from non-communicable diseases, appear to respond only to significantly larger government expenditures. An efficient allocation of scarce government dollars for health care would target the “low-hanging fruit” first, striving for reductions in the factors that influence infant and childhood mortality. As a country’s wealth improves, it could have more money available to target more expensive and difficult health-care challenges, such as non-communicable diseases.
At the same time, many of the factors that promote the health and life-expectancy of a population may be only indirectly related to health care. Improved economic status, education, nutrition, and environmental quality may do more to improve the health care outcomes of a country than specific spending on health care. Government spending in these areas could result in significant improvements in important health outcomes, while not being reflected in the direct expenditures on health care per se.