It's the Prices, Stupid: Why The United States Is So Different From Other Countries

Gerard F. Anderson, Uwe E. Reinhardt, Peter S. Hussey, Varduhi Petrosyan

Health Affairs. 2003;22(3) 

In This Article

Capacity and Utilization

Table 3 and Table 4 present selected data on the supply side of the health systems in the OECD. There is considerable variation in the composition of the supply side and in reported utilization rates. A limitation of these data, of course, is that they mask important differences in the specialty composition of the physician supply and in the content of crude utilization rates, such as "physician visits," "hospital admissions," and "acute care hospital days."

The general picture that emerges from Table 3 is that the number of physicians per 1,000 population (physican density) increased in most of the OECD countries during the 1990s. As the Table also shows, however, there are some exceptions to these general trends. In both Canada and Sweden physician growth was limited to population growth during the 1990s. In the United States medical school enrollment has been essentially constant since 1980. The observed increase in the number of physicians has mostly come from physicians who immigrated to the United States following medical education in other countries.[11]

Richard Cooper and colleagues have argued that a common driver of physician density in all industrialized countries has been economic growth, represented by GDP per capita. The authors observe that within OECD countries, GDP and the number of physicians per capita are highly correlated.[12] However, countries with higher GDP per capita are not more likely to have more physicians per capita than are countries with low GDP per capita.[13] This suggests the importance of factors unrelated to GDP in determining physician supply differences. Several commentators have observed that a causal link between GDP and physician supply may be overly simplistic.[14]

While many OECD countries perceive a nurse shortage, the actual number of nurses varies considerably across the OECD countries ( Table 3 ).[15] The number of nurses per 1,000 population (nurse density) ranged from 1.1 in Turkey and Mexico to 14.7 in Finland, and the number of nurses per acute care hospital bed ranged from 0.3 in Turkey to 1.5 in Norway. The United States ranks higher than the OECD median on both measures, although several of the European countries report a higher nurse density than does the United States.

Some researchers have contended that as a population ages, the demand for nurses will grow rapidly.[16] The OECD data show that there is no significant correlation between the percentage of population age sixty-five and older and the number of practicing nurses per 1,000 population.[17] However, there is a significant positive correlation between the growth rate of the percentage of population age sixty-five and older and the growth rate of the number of practicing nurses per capita between 1990 and 2000.[18]

Most of the OECD nations greatly reduced the number of acute care hospital beds, the average length of acute care hospital stay, and the number of acute care hospital days per capita during the 1990s ( Table 4 ). Turkey and Korea, however, increased their systems bed capacity, and the United Kingdom increased its average length of hospital stay slightly.

The German and Swiss health systems appear particularly well endowed with physicians and acute care hospital beds compared with the United States. The two countries rank much higher than the United States does on hospital admissions per capita, average length-of-stay, and acute care beds per capita. The average cost per hospital admission and per patient day in these countries must be considerably lower than the comparable U.S. number, however, because both countries spend considerably less per capita and as a percentage of GDP on hospital care than the United States does. The average U.S. expenditure per hospital day was $1,850 in 1999 three times the OECD median.[19]

Explanations for differences. There are several plausible explanations for this difference. First, the inputs used for providing hospital care in the United States - health care workers' salaries, medical equipment, and pharmaceutical and other supplies - are more expensive than in other countries. Available OECD data show that health care workers' salaries are higher in the United States than in other countries.[20] Second, the average U.S. hospital stay could be more service-intensive than it is elsewhere. While this may be true, it should be noted that the average length-of-stay and number of admissions per capita in the United States are only slightly below the OECD median. Third, the U.S. health system could be less efficient in some ways than are those of other countries. The highly fragmented and complex U.S. payment system, for example, requires more administrative personnel in hospitals than would be needed in countries with simpler payment systems.[21] Several comparisons of hospital care in the United States with care in other countries, most commonly Canada, have shown that all of these possibilities may be true: U.S. hospital services are more expensive, patients are treated more intensively, and hospitals may be less efficient.[22]

U.S.-Canada comparisons. Some in the United States believe that Canada is rationing health care by placing tight constraints on capacity and waiting lists. That impression is reinforced annually by the annual waiting list survey of Canada's Fraser Institute.[23] Table 4 shows that hospital admissions per capita, indeed, were lower in Canada than in the United States in 2000. Remarkably, however, Canada actually had a higher acute care bed density than did the United States and also reported a greater number of acute care hospital days per capita. The explanation for this seeming paradox could be the much longer average length of hospital stay in Canada. In both 1990 and 1999 the Canadian length-of-stay exceeded the comparable U.S. numbers by about 20 percent. To the extent that bed capacity is a binding constraint in Canada, further reductions in average lengths-of-stay could help to relax that constraint.

Medical technology. Hospital beds and health professionals are, of course, not the only binding constraints on a health system's capacity. Just as constraining, and possibly more so, can be the availablity of advanced medical technology. As shown in Table 5 , Canada has far fewer computed tomography (CT) and magnetic resonance imaging (MRI) scanners per capita than the United States does. Indeed, Canada's endowment with this type of equipment lies considerably below the OECD median, although Canada's is the fifth most expensive health system in the OECD.[24] As is further shown in Table 5 , Canada's health system also delivers far fewer highly sophisticated procedures than does the U.S. system. For example, the U.S. system delivers four times as many coronary angioplasties per capita and about twice the number of kidney dialyses. These data, of course, do not provide insight on the medical necessity of these procedures.

Quite remarkable, and inviting further research, is the extraordinarily high endowment of Japan's health'system with CT and MRI scanners and its relatively high use of dialysis. These numbers are all the more remarkable because Japan's health system is among the least expensive in the OECD.