Three decades ago, when the big switch from fee-for-service medicine to managed care began, there were two great expectations. The first was that preventive medicine would keep people healthier. Though progress occurred in some areas - smoking was reduced, for example - no great strides were made. The second was that competition would continue to drive down costs. That didn't happen.

In an article in the Harvard Business Review, professors Michael Porter of Harvard and Elizabeth Teisberg of the University of Virginia summed up that failure in a sentence. Competition is being waged over the wrong goal. The great battle in health care is over "who pays" and not, as it is in most arenas, over who provides the best value for the dollar. Until the goal changes to increasing value instead of merely dividing it among players, health-care quality will continue to decline, costs to rise and the ranks of the uninsured to grow.

The authors have several prescriptions for change. Taking them will be necessary no matter who winds up paying for health care - employers, individuals or the government.

One of them will strike many players in the system as radical: an end to the volume discounts that allow large employers and groups to demand lower prices from providers. The cost to treat a given problem is the same whether the patient belongs to a small group, large group or no group at all. So a discount given one patient doesn't reduce costs. It just shifts them to someone else and adds to costs when players scrap over the bill.

To make truly informed choices, patients and employers need far more accurate information than they are now given. That information should include prices that are clear, understandable and comparable. That's not the case now.

Information about a provider or health plan's experience and success rate should be specific. It should be easy to compare how good Hospital A is at treating and managing a given disease or condition compared to Hospital B. Publication of such information will lead providers with poor track records to improve or get out of the business of treating particular diseases. Similar information should allow buyers to compare health plans and determine whether Insurer A does a better job than Insurer B in managing diabetes, for example.

The authors cite the pioneering work of Dartmouth Medical School's John Wennberg. He found that although the annual cost of treating a Medicare enrollee ranged from $3,500 to $8,500, there was no correlation between higher costs and improved medical outcomes. The fault lies with the current system, which rewards patients for seeking care from within a narrow geographic network. That reduces the competition that would normally lead to improved service.

To increase efficiency and achieve the best results, providers should focus on the things they do best and most often. Competition should be over who can best treat a given problem. Patients should be given the information they need to pick the provider with the best track record and be free to seek treatment there without paying a penalty.

Since a better outcome will mean lower costs in the long run - fewer subsequent hospitalizations, for example -value will have been added to the system, not subtracted through haggling, paperwork and lengthy battles for referrals.

The authors believe that if buyers are given the information they need to compare insurers and providers on the basis of long-term value rather than this year's price, they will make the same choice all consumers do. They will pick the product that gives them the most for their money. We agree.

And when all the players in the health care system compete on that basis, the mad increase in prices will slow while quality increases.