Evidence can be found from particular Member States of the effectiveness of particular measures which were taken. The cost containment measures have been broadly effective as shown in Table 7. Total expenditure on health care (public and private) grew much less in relation to the Gross Domestic Product in the period 1981 to 1991 than in the previous decade. In Denmark the proportion fell. The exceptions were Italy and Greece. Comparing 1986 to 1991, the proportion fell in three Member States - Germany, Greece and Ireland.

TABLE 7. - Health expenditure as a proportion of gross domestic product

Source: OECD, 1993.

Expenditure on pharmaceuticals grew much more than total health expenditure between 1980 and 1990 and cost containment measures have not been very effective (see Table 8), although the pharmaceutical sector is the «usual suspect» in almost every cost containment initiative.

TABLE 8. - Pharmaceutical expenditure as a proportion of the Gross Domestic Product

Source: OECD, 1993.

Controls by cash-limited budget, target budgets, or quotas over the whole health sector or important parts of it and manpower limits can clearly be made to work. Budgets for hospitals have accelerated reductions in length of stay and budgets for doctors' earnings in Germany have reduced them. Increases in cost-sharing can reduce demand, although, if relatively high, as in France, supplementary insurance may develop, thus negating some of the effects on demand and increasing total expenditure. Changes in the relative value scale in Belgium finally stabilised expenditure on diagnosis tests as have expenditure ceilings in France and Germany.

Increasing cost-sharing is likely to achieve results on a «once and for all» basis.

Figure 1 shows the number of items dispensed by chemists in the UK in relation to NHS prescription charges. It is clear that although there were fifteen increases in nearly fourteen years pharmaceutical consumption increased considerably. The introduction of a limited list in 1985 does not seem to have reversed the trend and produced only a one-off effect in 1986.

The explanation could lie in the fact that only one in six items dispensed on the NHS is paid for by the patient and that only approximately 40 per cent of the population are not exempt from paying for their own prescriptions.

The same trend is illustrated in Figure 2 for France where several cost containment measures and increases in co-payments failed to curb escalating expenditure (LE PEN, 1994).

It is often difficult to distinguish the quantitative effect of particular measures because of other changes happening at the same time. An example is the reference price system for pharmaceuticals in Germany. Another example is that of co-payment evasion in Spain.

Pharmaceutical companies increased the prices of products not yet affected by the reference price system. Between 1991 and 1992 drug prices subject to reference prices decreased by 1.5 per cent while drug prices in the segment free of reference prices increased by 4.1 per cent.

Reference prices were set, on average at 30 per cent below the previous price of the brand name products. The expected boom of the market for genetics did not occur. In terms of sales revenue, the annual growth of 2 per cent of the generic market prior to 1988 Health Care Reform Act slowed down to about 1 per cent annually since. This may be explained by new product-life cycle strategies of R&D based companies, above all, the attempt to reduce price drastically after patent expire to create barriers to market entry. Companies also reduced prices of unpatented products because patients were unwilling to pay for the difference above the reference price level.

In addition, a large part of the market was inaccessible for reference pricing, allowing pharmaceutical companies to pursue effective «loop-hole». strategies. In 1992, the patent-free share of the sickness-fund market comprised some 300 drugs with a value of £3.84 bn which represents 39.3 per cent of the total market.

FIGURE 1. - NHS Prescription charges and items dispensed by chemists *. U.K.

Note:* Including appliance contractors.
Source: OFFICE OF HEALTH ECONOMICS, 1995.

FIGURE 2. - 1976-1993: Failure of cost-containment policies

Source: LE PEN, 1994

Reference prices did not prevent increases in volume in all market segments and in some cases physicians prescribed expensive patented products and ignored cheaper alternatives. Table 9 presents the annual increases in pharmaceutical expenditure, volume of prescriptions and price per prescription between 1988 and 1993 as a percentage over the previous year. According to Schwabe and Paffrath (1994), a third component, the structural component (particularly the changes in the package size of prescription drugs) also contributed to increases in pharmaceutical expenditure. This led to the link of co-payments with package sizes.

The examination of profiles of doctors' work and prescriptions only seems to have limited effect but this may depend on what sanctions are applied and how often they are imposed.

Initial evidence suggests that contrary to expectations, fundholding in the UK has made no difference in the referral rate.

The same happened in Germany, although the overall drug budget seems to be very effective in its first year of operation. The overall drug budget for 1993 set at £9.56 bn, fell sort of £880 m or 9 per cent. These savings are due both to a reduced number of prescriptions compared to 1992 (-10.4 per cent) and a reduced value of each prescription (-4.6 per cent).

TABLE 9. - Germany: annual percentage increases of total pharmaceutical expenditure, volume of prescriptions, cost per prescription and «structural component» between 1988 and 1993

Source: SCHWABE and PAFFRATH, 1994.

The therapeutic categories which were mostly affected by the introduction of the overall budget were peripheral vasodilators, analgesics and antirheumatic products. Other categories such as antidiabetics and antibiotics or ACE-inhibitors were not affected.

However, as in the UK, referrals to other specialists increased by 9 per cent and referrals to hospitals where drug budgets do not apply increased by 10 per cent. Schulenburg et al. (1993) calculated that these alternative strategies incur additional direct costs of £520 m to the sickness funds plus additional indirect costs (loss of productivity) of £600 m.

In Spain in 1992 the number of prescription per active worker was 7.1 and that of a pensioner 39.3 which corresponds to 3,275 prescriptions per month. Per capita expenditure for pharmaceuticals for the active workers and the pensioners was 5,820 and 49,272 pesetas respectively. Pharmaceutical expenditure for pensioners is 8.5 times higher than that of the active workers. This is a very high ratio, considerably higher than in other countries (especially if we take into account that many pensioners are not over 65) and an explanation may be fraudulent practices involving the use of pensioner prescriptions by persons who are not exempt from co-payments.

In Italy the combination of the «average price» system with the new positive list and the exclusion from reimbursement of more than 50 per cent of the products, had as a result a considerable reduction in pharmaceutical expenditure (-15.1 per cent between 1992 and 1993). However, it is still early to assess any long term effects because the political situation in the country is not yet stable and the pharmaceutical industry has not yet recovered from the scandals revealed in the last two years.

Price controls may also lead to perverse incentives through the introduction on the market of non innovative products.

In some countries pharmaceutical companies have been trying to bypass strict pricing controls for old products and competition for off patented products by launching new products, which are not necessarily innovative.

Table 10 shows the market shares of new products on the market of six EU Member States between 1991 and 1993. It is clear that the market shares of real novel products are low in all the countries examined and this may reflect the low number of innovative products introduced on the relevant markets. However, market shares of new products which do not offer real improvements are very high in Spain, Italy and Germany were strict price control systems or reference price are in force.

TABLE 10. - Impact of new products on total pharmaceutical spending (Products introduced in last three years in percentage of market share)

Note:
A: All New Products (%)
B: New Chemical Entities (%)
C: Novel Products (%)

Source: CUENI, 1994.

From the above it is clear that specific measures may create regulatory bottle necks which can easily shift costs to other sectors and categories.

The obvious way ahead is for EU Member States which have not yet done so to consider introducing those measures which have proved effective in other Member States. It is clear than some types of action are effective such as budget controls, if they are rigorously enforced, and in some cases backed up by manpower controls. Cost-sharing can transfer costs from the public sector to the private sector and restrain some costs in total, providing this is not counteracted by extensive private insurance of these co-payments. If such insurance was forbidden or discouraged by the removal of any tax concessions, cost-sharing measures would be more uniformly effective in achieving their objectives. But if they were substantial or if attempts to restrict the effects on the poor are not very effective, de-insuring in this way can have damaging effects on equity and increase inequalities in access to health care.

More Member States could encourage generic prescribing in a variety of ways and more could have positive lists. It is notable that most of the countries without the latter are major exporters of drugs and, presumably, fear repercussions on their exports. Some Member States have hesitated to go far in rationalising their hospital stock or adopting effective ways of limiting the proliferation of expensive medical equipment, outside as well as inside hospitals. Only preliminary steps have so far been taken in making a reality of a single market for drugs. This should lower the prices of drugs in some Member States and probably increase them in others.

In some Member States, costs could be saved by increasing the number of general practitioners at the expense of specialists and establishing a pattern of referral, as specialists are more likely to use expensive specialised services when it is not strictly necessary. Germany has taken steps to go down this road. Moreover, it is increasingly accepted that an excess of doctors in health insurance practice leads to an excess of costs.

At present, sufficient information is lacking on the long term effects of many of the actions taken by some Member States. The difficulty is that a measure taken in one Member State is often quickly followed by another before there is time to see the long term effects of the first measure.