* Originally published as "La santé comparée en Europe et en Amérique," Données sociales, INSEE, 1999 edition.

** Gérard Badeyan, Daniel Foulon, Jacqueline Gottely and Patrick Pauriche are part of SESI at the Ministry of Employment and Solidarity.

At the beginning of the century, inhabitants of what was known as the "developed" world lived on average less than fifty years. By the end of the century, they have a life expectancy approaching 80 years. If the health and social security systems do not in themselves alone explain this fantastic progress in life span, they are nonetheless the framework and provide its sustainability.

At a time when we are all anxiously seeking to safeguard, through inevitable change and adjustment, the essentials of our health and social protection systems, the debate conducted by our neighbours and the solutions they plan are of great interest. The comparative investigation reported in this article (see box) covers a fifteen-year period (1980-1994) punctuated by major events in the health field: the rapid development of medical genetics; the integration of scientific and technical progress, especially in the medical imaging and functional exploration field; changes in medical practice, dealt with in part in this article under the heading "day surgery"; but the appearance of new diseases too, with AIDS, hepatitis C and illnesses linked to the environment, especially allergies.

The countries chosen for the comparison (Germany, Canada with Quebec, the United States, France, the United Kingdom), taken from among the G7 countries, are all wealthy and economically developed. They have quality medicine (the recipients of 114 of the 158 Nobel Prizes for Medicine awarded from the outset to 1994 were from these five countries) and wide social protection. This choice allows access to reliable information and the general characteristics of the countries are sufficiently close to enable the analysis to be better centred on the indigenous components of health.

Alongside the United States, inspired by a liberal philosophy, we find countries where health insurance takes the preponderant place (Canada, Quebec, France, Germany) and one country where healthcare costs are guaranteed for all residents by the state, the United Kingdom. Beyond these general principles, the nature of the country has consequences on the organisation of the health system: the province in Canada, the State in the federation of the United States, the Land in Germany have, to varying degrees, a significant responsibility in organising and running the health system. Access to the health system can be very free as is the case in France or Quebec; it is more supervised in Germany (where illness first has to go via a doctor in the outpatient sector to access the hospital) and in the United Kingdom (where patients are registered with a list of general practitioners under contract to the National Health Service).

In Canada, as in Quebec, hospital insurance and illness insurance form the keystones of the health system. There is no patient contribution within these two insurance schemes.

In Germany, the system is structured around occupation based health insurance schemes. The insurance is compulsory for the most modest categories. Other people have the choice between more than six hundred funds. The patient contribution is fairly widespread (for medicine and hospital), even though there are exemptions.

In France, finance is basically provided by Social Security which incorporates about fifteen health insurance schemes offering different levels of repayment. Most people have additional cover from insurance funds and insurance companies. The patient contribution and excesses compared to the rate recognised by the Social Security are fairly widespread.

In the United States, old people are covered by Medicare and the poorest by Medicaid. Americans who work in a company of a certain size are covered by health insurance financed both by themselves and by the employer. A significant number of Americans (about 15%) live without health insurance.

In the United Kingdom, the system is organised round the National health Service (NHS) open to all and financed by taxes. However, the patient contribution is not unusual in this system.

By various means and with varying degrees of success, most countries have tried to control their health spending. This is because there is a massive trend for spending to increase, due in particular to ageing of the population, technical advances and a widening of the services available. It seems that the existence of a solid system of guiding patients towards the different healthcare sectors and the search for alternatives to remuneration for care enable costs to be contained.

In the United States HMO (Health Maintenance Organisation) formulas have developed and other variants known under the generic name of managed care. These involve private companies that, for a set price, insure all health services relative to a group of people.

In the United Kingdom, the NHS has been restructured to make it less dependent on the state and to introduce competition within the internal market.

In Canada, the provinces’ contribution to financing health spending has increased whereas the federal government share has reduced. Provinces have had to rationalise the system and control costs by reducing the time patients remain in hospital, a measure made easier by improvements in therapy techniques. This is especially the case in Quebec.

Germany has reformed the organisation of health insurance by introducing competition between health funds. It has also set restricting budgets, frozen prices and raised patient contributions.

France, after introducing the overall hospital grant in 1984-1985 and implementing quantified national targets for most of the health professions in the early nineties, is engaged in the November 1995 reform. This established regional hospital agencies, regional unions of health insurance funds and provided for Parliament to vote a limited amount of expenditure under the Social Security finance law.

For all of the countries investigated, in 1994 we see a ratio of 1 to 3 between per capita health expenditure expressed in US dollars and in parity of purchasing power (i.e. with a currency conversion that eliminates price level differences between the countries). The United Kingdom spends about $1,200 per capita on health. This low level of expenditure is explained by the preponderance of the role of public institutions, especially within the National Health Service. On the other hand, the liberal organisation of the health system in the United States leads to a high level of spending of about $3,600 per capita. France is in a position very close to that of Germany and Quebec with about $1,870 per capita in 1994. Between 1980 and 1994, the United States, where spending levels were already the highest, also experienced the highest level rate of increase: expenditure there has increased 3.3 times between the two dates. Measured as a proportion of GDP allocated to health spending, the differences are also substantial. The United States in 1994 spent 13.5% of GDP to health, France, Canada and Germany about 10%, and lastly the United Kingdom a mere 6.9% (figure 1).

The different organisation of health systems arises in part from the commitment of the state and social insurance at levels varying according to country. This commitment can be measured by the public share of health expenditure. In the United Kingdom this share is the largest: about 84% in 1994. In France, Canada and Germany, it is slightly lower at 75%. In the United States it does not exceed 45% and basically equals the cost of the Medicare and Medicaid programmes paid by the federal government. Between 1980 and 1994, the general trend is a reduction in the public share of health expenditure, except in the United States where it increased about two points over the period.

If we look closer into the structure of expenditure on health services and assets, we see that hospital spending everywhere represents the largest share of expenditure, even if its share has fallen between 1980 and 1994 in most countries. In 1994, the proportion varied from 45% in Germany to 60% in Quebec. In France, it came to 51%. Outpatient care varied between one quarter of spending in Quebec to 37% in the United States (about 30% in France). Lastly, medicines represent about 10% of spending in the United States, 19% in France and 23% in Germany.

For most healthcare professions, the density of practising personnel relative to the population varies in a ratio of 1 to 2 (figure 2). The practising personnel available are structured differently depending on the organisation of the health system. In Germany and France, the density of doctors and surgeons is high whereas the United States and Canada are distinguished by the size of their paramedical professions. The unequal distribution goes back to differing distributions in the field of skills between doctors and nurses according to country, as well as to the respective roles of in-patient and out-patient departments. The United States is unique with its very low proportion of general practitioners among all doctors (13% in 1994).

Control of the range of healthcare tends to be broadening throughout. Whatever the profession and density of practitioners, numbers employed in healthcare increased slower between 1978 and 1987 than over the previous period (1980-1987), with the exception of the United States. Numbers actually fell in the United Kingdom. This slow-down more affects general practitioners than specialists. As an example, the number of doctors being trained was cut by half in France between 1980 and 1994. Other types of control can be seen, such as the emigration of Canadian graduates to the United States or, conversely, the recruitment in the United States of doctors trained abroad. Similarly, a proportion of doctors in Britain originated abroad. Lastly, the proportion of women in medicine has increased everywhere, but remains higher in Europe where it peaks at 34% in Germany.

Whatever the country, qualified nurses are the most numerous profession and their proportion among all healthcare workers is increasing in most countries. The density of qualified nurses and midwives varies from single to double, from 45.9 per 10,000 population in Germany to 90.6 in Canada. The increase in nurses was substantial everywhere from 1980 to 1987, although lower than that for doctors, except in the United States, which is the sole country where the number of nurses continued to increase strongly from 1987 to 1994 (+3.3% per annum). In other countries the increase has slowed down more or less sharply, even becoming negative in the United Kingdom.

"Nnursing assistants" work, within their sphere of expertise, under the responsibility of a nurse, basically in hospital. Their density is greatest in France and the United States (35.4 and 35.1 respectively per 10,000 population in 1994). Germany appears atypical with a very low density (8.5 per 10,000). In part, this is explained by the presence of many volunteers in the German hospital system, notably people doing their national service. If the density of auxiliary nurses reduced or remained stable in most countries, it increased greatly in France between 1987 and 1994 because of measures to redistribute some nursing work to nursing auxiliaries after lightened training.

If the United Kingdom is excepted, the amount of healthcare measured by the number of general practitioner and specialist consultations is greatest in France (2,720 in 1994 against 1,674 in the United States). The general increase in medical density on the one side and changes in medical consumption on the other have not had the same consequences in terms of medical activity in all the countries.

This activity was relatively stable in Canada throughout the period. In Quebec, it fell 14% between 1980 and 1994. In the United States, it fluctuated according to the period. France distinguishes itself from other countries by a constant increase in the activity of doctors. If in the United Kingdom the number of consultations per general practitioner is particularly high, France comes top for the activity of specialists (figure 3). In all countries, general practitioners provide more consultations than specialists.

Main source of health expenditure, the hospital has found itself right at the centre of the mechanisms implemented with a view to limiting costs. The trend, towards greater rationalisation or even greater "profitability" of hospital activity is general. Hence the number of beds per capita fell in all countries between 1980 and 1994, a probable consequence of the restructuring and closures that took place. This is equally true for countries where the availability of beds is high, as in Germany or France, as it is for the United Kingdom which still has the lowest number of beds per capita (figure 4). In 1994, short stay bed occupancy was much higher in the United Kingdom and far less high in the United States than elsewhere, whereas these rates varied little from one country to another in 1984.

Hospital admission rates measure the amount hospitals are used. Their increase in Europe and decrease in the North America has resulted in a clear difference between the two continents. The number of days of short term hospital stays per capita has decreased sharply in most countries, a little more slowly in Quebec and very little in Great Britain where, it is true, it was the lowest in 1980. The United States is a case apart, with a relatively low number of days of hospital stay relative to the population: one third of that for Germany, half that for France. The average stay for short term physical care has reduced everywhere, more in Europe than in America. Germany is unique with an especially long average stay (11.5 days in 1994 against a minimum of 5.2 days in Great Britain). Lastly, the number of staff per bed divides the six countries into three sub-groups, with a lower rate of staffing in Germany and France, intermediate in Canada and Quebec and highest in Great Britain and the United States.

The frequency of "day", or out-patient, surgery is a fairly relevant measurement of technological development and medical practices. Met with much more frequently in North America where nearly half of all surgical procedures are now out-patient whereas its frequency barely exceeds 10% in Germany.

"High technology" equipment (one of the measurements of the moderness of medical practice) is more widespread in the United States, then in Germany, than in other countries. France is unusual however in having much more cardiac catheter equipment than its four partners.

All of the countries investigated are enjoying higher life expectancy at birth. In 1994, it varied for men from 72.4 years in the United States to 75.1 in Canada, and for women from 79 years in the United States to 81.8 in France (figure 6). From 1980 to 1994, life expectancy at birth increased everywhere and for both sexes. There are more male deaths, whatever the age or country. The difference in life expectancy between men and women has, however, reduced in over the period in North America and the United Kingdom. The greatest difference is seen in France (8.1 years), as was already the case in 1980.

One of the major facts of demographic change seen over the period is unquestionably the increase in life expectancy in old age. The increase is proportionally two to three times greater than increases in life expectancy at birth achieved over the same time lapse. From 1980 to 1994, the greatest increase was seen in France where life expectancy at 65 rose from 14 years to 16.2 years for men and from 18.2 years to 20.6 years for women. At the opposite end of the scale is the United States: life expectancy at 65 there went from 14.1 years to 15.5 years for men, and from 18.3 years to 19 years for women.

The fall in infant mortality continued and the progress achieved has been spectacular, especially in countries where it was highest. In 1994, infant mortality rates were close in all countries except the United States. Thus in 1994, the infant mortality rate varied from 5.6 deaths of infants under one year old per thousand live births in Germany and Quebec to 6.3 in Canada, but reached 8 per thousand in the United States.

Early neonatal mortality, that for infants dying at under seven days, represents about half of all infant mortality. Germany has made the greatest progress in reducing it, on the other hand, it has reduced little in Canada and the United Kingdom between 1987 and 1994. Thus in 1994, the rates run from 2.3 per thousand live births in France to 4.2 in the United States.

The allocation of deaths into broad groups reflects, for the six countries investigated, the situation that currently prevails in industrial societies: the pre-eminence of deaths due to heart disease and cancer. In 1994, two large groups of causes represented 73% of deaths in Germany, 69% in the United Kingdom, 65% to 66% in North America and only 60% in France which thus distinguished itself from its European neighbours (figure 7).

From 1980 to 1994, we see a substantial decrease in deaths from heart disease, from 30% in the United States and Germany to 42% in Quebec. Comparatively, death from tumours decreased very little. Among men, it decreased in all countries, from 1% in Canada to 7% for the United Kingdom. Among women, it decreased in the three European countries, from 2% in the United Kingdom to 8% in France, yet increased in North America.

If deaths due to injuries decreased everywhere substantially, from 20% (United States) to 35% (Germany), deaths due to respiratory diseases changed in sharply differing ways: over the whole 1980-1994 period we see a reduction of around 20% in the three European countries against an increase in North America, from 3% in Canada to 20% in the United States, an increase due essentially to deaths among women.

Within this whole picture, France enjoys a very favourable position in cardiovascular disease, the opposite of Germany. For cancers, it shows the particular characteristic of having the highest death rate among men and the lowest among women. France is also unique in that injury is its third largest cause of death (8.7%) owing to the large number of suicides and traffic accident deaths. Third place is occupied by respiratory diseases in all five other countries. For this type of illness, the United Kingdom occupies a separate place with rates at least one and a half times greater than those in other countries and, for women, up to three times greater than those in France.

Lastly, the importance if AIDS among causes of death varies enormously from one country to another: in 1994, we find a ratio from 1 to 16 between male death rates and from 1 to 23 for female death rates between the United Kingdom and the United States.

Beyond the analysis of all deaths, it is fundamental in terms of public health to investigate deaths occurring at what is considered to be an early age. If we than analyse deaths occurring before the age of 70 (this being the agreed age limit), by weighting deaths occurring at the youngest ages (see box), we are able to make the following assessment. All injuries together represented the main cause of premature death among men, in 1994, except in the United Kingdom. Premature death due to injury and cardiovascular disease are both highest in the United States. In France, this is the case for all tumours, and in Quebec for the specific case of lung cancer. Everywhere we see the increasing trend of premature death by suicide among men, except in Germany.

This cause of death is more important in Quebec than in the other countries. Furthermore, premature death from AIDS is far greater in the United States.

Among women, all tumours largely dominate premature death, breast cancer being the most frequent. However the increasing importance of lung cancers should be emphasised, especially in Quebec and Canada.

Thus between countries there are convergences or disparities that stand out for researchers and decision-makers. If a few general trends do appear, some of the differences between countries have widened and others, on the contrary, have reduced. How does the performance of the six systems compare? Is one better than the others? Without answering these questions directly, some findings nonetheless stand out: the most highly state controlled system, the United Kingdom’s, is also the least costly; the most liberal system, in the United States, is the most costly and the least egalitarian in terms of access to healthcare, but it is also the most modern and high tech in its "rich" part. Lastly, performance measured in absolute terms, life expectancy, firmly places the French system at the head of the pack where women are concerned.

Indicateurs socio-sanitaires, comparaisons internationales, évolution 1980-1994. SESI, Ministry of Employment and Solidarity (France) and DIRIS, Ministry of Health and Social Services (Quebec), La Documentation française, 1998.

This article is a summary of the Indicateurs socio-sanitaires, comparaisons internationales, évolution 1980-1994¹ report produced jointly by experts from DIRIS (statistical office) in the Ministry of Health and Social Services, Quebec, and SESI (statistics, reports and information systems department) of the Ministry of Employment and Solidarity, France. Their German, American, British and Canadian opposite numbers were directly associated in drafting the report.

Thirteen procedures were investigated which occupy a very broad spectrum in terms of pathologies treated: from the removal of tonsils to kidney transplant (figure 5). The analysis covers the year 1993.

Divergence in the frequency of some procedures between the six countries is primarily explained more by differences in the practices of health professionals than by differences in the state of health of the populations investigated. Two countries set themselves apart by the extreme positions they occupy: the United States and England. The United States is typified by the generally very high rate of surgical procedures: in seven cases out of thirteen, the frequency of surgery was highest and only once was the lowest. At the opposite end of the scale, England is set apart by rates that are generally very low. France is uniquely situated with rates sometimes very high and sometimes amongst the lowest.

So with 355 appendectomies per 100,000 population, surgeons operate 2.5 times more frequently in France than in the United States which comes in second place, and 3.6 times more frequently than in England where the operation is performed least often. The differences lead us to query the relevance of some procedures.

Another procedure is also very frequent in France: total hip replacement which has the purpose of giving the patient back the ability to move without pain (119 operations per 100,000 population, or 3.4 times more than in Quebec).

On the other hand, relatively few heart by-pass operations are performed in France (an operation intended to re-establish better circulation of the blood by by-passing the obstructed vessel): the procedure is performed 4.2 times more in the United States and 2.1 times more in Canada. This particular situation can be related to the state of health of the population, because France is typified by a low rate of heart disease (which includes in particular myocardial infarction and angina pectoris).

In Quebec, the rates for lobectomies of the lung (the removal of a lobe) and pneumonectomies (the removal of an entire lung) are especially high. This can be explained by the rate of deaths from lung cancer for which Quebec holds the sad record among the countries monitored. This interpretation has a limited scope however because the ratio between the two figures is not seen in other countries.

In the opposite direction, it is Quebec that has the lowest the rate for total mastectomies (total removal of the breast).

Premature death is analysed through the indicator of potential years of life lost. This gives an estimate of the total number of years’ life lost before a limit age set in this investigation at 70. For a given cause, the number of years’ life lost is a function of the number of deaths and the average age of the persons dying. The advantage in using this indicator is that it gives more weight to deaths occurring at a young age.

Figure 2 - Density of health professions in 1994

1. Number of qualified new doctors per 10,000 population.
2. General practitioners and specialists.

Source: Indicateurs socio-sanitaires, comparaisons internationales, évolution 1980-1994
SESI and DIRIS, La Documentation française, 1998

Figure 3 - Number of consultations per doctor, per capita in 1994

1. This number comes from a household survey. It is probably overestimated
for two reasons, first, individuals may list consultations outside the survey
period, and second, there may be confusion between seeing the doctor in person
and seeing a nurse in the practice.
2. Great Britain.
3. Number of consultations per child under fifteen.

Figure 4 - Indicators of hospital services for short term physical care

1. Hospital admission rate: number of admissions in relation to country’s population (as %).
2. Bed occupancy rate: ratio between the number of days in hospital and the theoretical
total capacity in days (number of beds x number of days open in the year).
3. For Germany, this is the change for 1990-1994, given as an indication.

Figure 5 - Standardised rates per 100,000 population for particular surgical procedures in 1993

1. Data not available for the whole United Kingdom.
2. 1994 data other than for kidney transplants (1993).
3. Ratio between the highest rate and the lowest rate.
4. Rate per 100 births.

Figure 6 - Some demographic indicators in 1994

1. Rate of infantile mortality: number of deaths of children under 1 year old per 1,000 live births.
2. Rate of early neonatal mortality: number of deaths of children under 7 days per 1,000 live births.

Figure 7 - The main causes of death in 1994