Death. Taxes. Medicare.

We are destined to experience all of them, though not in that order. What the three have in common is that each one, by itself, is more than most of us can think about. They make your head hurt.

The unfunded liabilities of Medicare, according to the latest generational accounting figures from economists Jagadeesh Gokhale and Kent Smetters, now amount to $35.5 trillion. That's about 5 times the $7.2 trillion unfunded liability of Social Security. It's also about 10 times the Treasury debt held by the public that we worry so much about.

Worse, the Medicare liability is growing with the speed of a nasty cancer. It will increase another $1 trillion by next year's Presidential election, dwarfing the recognized federal deficit. Then it will grow another $5 trillion by the Presidential election of 2008.

This is a bet-the-country sized problem. Whether you are 10 or 90, you will live through some part of it. Solve it and we get to continue being the great country we are. Fail to solve it and we go on the scrap heap of history in one generation. As a nation, our greatest single need is to innovate in the methods, delivery, and pricing of health care.

George B. Bennett may have a solution for a chunk of the puzzle. Mr. Bennett is the President and CEO of Health Dialog in Boston, an entrepreneur, and a former management consultant. You can tell, within seconds of meeting him, that he is an enthusiastic--- no, I'd say passionate--- problem solver. This spring I spent three hours in his Boston office, listening to him explain what he calls "transitioning from managed care to collaborative care."

I asked him what the difference was.

"Instead of managing care, managed care was used to get discounts and, ultimately, to deny care," he said. "The managed care world is a world of rules." As an example, he pointed out that if a woman has fibroids managed care has a series of rules that determine whether she can have a hysterectomy.

"The new idea is to make sure the family has access to the information that is used to make the rules. The idea is to inform." He noted that many women are less eager to have the operation when they learn that most of the symptoms (and discomfort) of fibroids disappears after age 50.

"What you find is that more women opt out of a hysterectomy under collaborative care than would be denied care (a hysterectomy) under managed care."

In another example, he pointed out that men can take the PSA test for prostate cancer but there is a good chance of getting a fake positive or negative result. "There is, however, a 100 percent chance of urinary or potency problems if they have an operation. What most people don't understand is that for every man who dies of prostate cancer, 150 die with it," he said. He also pointed out that it takes, on average, 7 years for prostate cancer to metastasize and another 7 years to actually die from the cancer.

As you might expect, many men opt out of prostate related surgical procedures when they are fully informed of the odds and side effects. "It's not the efficiency of the surgeon, it's the understanding of the side effects," Mr. Bennett pointed out.

His Bible is an amazing healthcare database developed by Dr. John E. Wennberg at Dartmouth Medical School. It shows, Mr. Bennett says, that "Geography is destiny--- where you live has more impact on your treatment than your health." Sifting through the Medicare book of the Dartmouth Atlas of Health Care, for instance, I found these rather extraordinary facts:

•           Medicare reimbursements per capita varied by nearly 300 percent,                with the lowest ($3,074) going to the Lynchburg, Virginia hospital                referral region and the highest ($9,033) going to the McAllen, Texas                hospital region.

•           The odds of dying in a hospital ranged from under 20 percent in Tucson to                nearly 50 percent in Manhattan.

•            Referrals for coronary bypass operations were as high as 11.5 per                thousand Medicare enrollees in Redding, California but only 3.6 in                Santa Rosa, California.

Basically, both the costs and frequency of any medical treatment vary widely, indicating that medical practices are idiosyncratic rather than scientific. This same chaos is a window of opportunity for Mr. Bennett and his "information therapy" delivered by trained and impartial nurses rather than agents of cost control.

"It has something about human caring. It has something to do with trusted information. It has something to do with developing skills. A lot of the healthcare model is paternalistic. But this (collaborative care) gets people involved (in their own care).

"I think we will establish that 'information therapy' is as important as surgery and pharmaceuticals," he concludes.

And what has this got to do with costs and that $35.5 trillion liability?

Everything. Mr. Bennett points out that conservative treatment options--- the kind individuals tend to opt for when they are engaged and fully informed--- could lower Medicare expenditures by 30 percent. That's a nice saving on any amount of money. But on $35.5 trillion it's $10 trillion.