Death fits uneasily into medicare's bureaucratic pigeonholes.

More than 80 per cent of Canadians die in hospital, rather than at home, largely because hospital care is paid for while palliative home care usually is not. Medicare, then, has a built-in incentive that runs counter to the fundamental human desire to die in the dignity of one's home. This is not an efficient use of money, either. Hospitals are hugely expensive places to be.

It may seem like lunacy to expand medicare when it is already short of cash. But if the system is to remain healthy over the long term, it must be allowed to grow. Expanding medicare is akin to a young family buying a larger home than they can afford to furnish at the moment. They buy it because they expect their needs to change. No sense in buying a home only to vacate it in a year.

Roy Romanow touched on this need to keep up with change in a speech last June to the Canadian Nurses Association. "The main problem facing medicare today is that we designed it for another era," said the man who will present his major report on the health-care system next week. "When we first started debating medicare, most health problems were acute illnesses. However, now most health problems are chronic illnesses in an aging population."

Our health system provides poor care for chronic illnesses, he said. As a result, hospitals and long-term-care homes fill up with patients who have suffered preventable complications. Modernizing requires a more flexible approach.

To some extent, this is already happening. Government spending on home care has grown. Hospital spending as a share of health costs has dropped from 45 per cent in 1975 to 32 per cent today.

Nearly half of hospital spending is for patients 65 and over; shortening their stays, or serving them where they live, is in everyone's interest. It falls to the federal government to set national standards for the new areas, and to ensure that a basic level of service is available.

As a matter of principle, there should be no user fees for public health care. But if Canada is to extend the umbrella of coverage to new services such as pharmacare and home care, without going into debt, it may have to require a form of co-payment for those services, under certain conditions. It must not effectively bar the poor from being served; it must not penalize the working poor by making it more attractive to be on welfare; it must not create two tiers, by allowing for separate "luxury" and "economy" levels within the service provided; and it must help government stretch its dollars substantially further, after administration costs are taken into account.

Where the up-front costs of new services are prohibitively expensive, making reasonable demands on the system's "customers" is an acceptable tradeoff.

Home care: Medicare's perverse incentive keeps many patients in expensive hospital beds even when acute care is no longer needed, simply because they can't function on their own at home. The presence of these "bed-blockers" means that not every hospital bed gives good value for the money spent. Thankfully, the system is not so cold as simply to expel these patients, many of them elderly, onto the mercy of their relatives.

A limited home-care program could be created to enable patients to be released from acute hospital care with medicare coverage for a set period. Senator Michael Kirby proposed a 90-day program in his report last month, which sounds sensible. Governments already spend $3-billion on home care. The service took up 1.6 per cent of provincial health budgets in 1988-89; 10 years later it was 4.7 per cent.

But per capita provincial spending varies wildly, from $40 to $200. Creating a national program could ensure a basic standard of home care, and allow for growth within the country's means as the health system evolves.

About 160,000 Canadians a year need palliative care to ease their dying. Medicare dollars should be available for those patients at home. Changing employment-insurance provisions to permit workers leave of, say, six weeks to care for dying relatives, as Mr. Kirby suggested, is also worth considering.

Pharmacare: Patients in hospitals receive their drugs free. Once released, they must pay for those same drugs. Philosophically, the distinction makes little sense. On a practical level, however, it is difficult to eliminate.

Drugs are the fastest-growing portion of Canada's health-care system, up from $3.8-billion in 1985 to $15.5-billion in 2001 (and this does not include drugs dispensed in hospitals). First-dollar coverage would be a backbreaking expense. A 1997 study estimated that a fully funded national pharmacare program would increase public spending on prescription drugs by $4.3-billion. Saskatchewan tried a comprehensive drug2 plan a few years back but scaled it back when costs soared 30 per cent a year.

A national drug program should be based, as Mr. Kirby said, on the medicare principle that a health catastrophe should not destroy people financially. Currently, 100,000 Canadians spend more than $5,000 a year on prescription drugs.

We need to be cautious in committing ourselves to new programs. But Canada has to recognize how the health-care requirements of its population are changing. Unless the system responds to shifts in social needs and in technology, its claim on the public purse will weaken.