The current crisis and high public deficits have reopened the question of the sustainability of the welfare state at discussion . More and more voices are wondering about the threat that the recession may pose to a healthcare system that already costs 8.5 percent of Spain's GDP. In this context, co-payment has reappeared as one of the solutions (if not "the solution") in vogue. We propose to review some of the lessons we can learn from other countries that have already adopted this formula.

Co-payment is supported by a long list of international experiences from countries as diverse as Taiwan, Canada, Australia, Sweden or the USA, which endorse its contribution to reducing the use of health services and to reducing expense. But obsessing solely on the reduction of expense is an incomplete vision of the problem as it forgets its impact on the health of the population, which is the main goal of any healthcare system.

An implementation manager of copayment requires us to ask the following questions: First, what effect does this decrease in the use of health services have on the health of the population? Second, how do these effects change with patient characteristics? Third, which subject of health services (necessary and less necessary) falls more with the introduction of copayment? And finally, what form will copayment take?

International experiences can help us to answer some of these questions. The now classic study of the Rand Health Insurance Experiment in the United States, which between 1971 and 1982 assigned some 2,750 families to health insurance with different co-payments, stands out. Its results show, firstly, that a higher co-payment on average does not translate into poorer health for the population. Second, there are two exceptions: high-risk patients (hypertensive, with heart problems, etc.) and the poorest, who were disadvantaged by higher co-payments. Third, when patients pay part of the cost, they make many mistakes and do not know how to prioritize, so that they end up reducing their healthcare consumption of both valuable and less valuable treatments for their health. The only exception is emergency services, where copayment did not seem to reduce their use in truly urgent situations. This result is also corroborated by the most recent experience in Sweden.

We should remember that there are many types of co-payment: private insurance (which at final is a voluntary form of co-payment and could be boosted with tax breaks); co-payment for hospital catering, as a penalty for those who misuse health services, and so on.

With these lessons, we can conclude that, in the event of introducing a co-payment, we should incorporate a stop loss clause that limits the maximum expense that the patient can incur and, probably, this should vary with their income and health status. In addition, the worst thing we can do is to have a co-payment for seeing a family physician. There a reduction in the issue of visits may have an adverse role in prevention, while a co-payment for emergency use may be more advisable and even more so if it translates into subsequent reimbursement for all justified visits. The options are many, but the right solutions are fewer and looking at international experiences can help us find them.