May 30, 2001
STANFORD, CA-Statistical analysts frequently massage numbers in esoteric ways that only they can fully appreciate. But in a recent study at the Stanford Graduate School of Business, business operations researcher Stefanos Zenios has applied his expertise to a life-and-death problem. He has developed a model that is likely to make more donor organs available for kidney transplants and help thousands of kidney patients who might otherwise die.
For more than three years, Zenios has worked with medical colleagues on the problem of kidney transplantation. Each year, 40,000 patients with kidney failure wait for transplants from people who have designated their organs for donation after death. Many patients, though once healthy enough to withstand a transplant, watch their health decline during the often years-long wait for a kidney. Only 15,000 people get transplants every year because of the organ shortage.
Of course, patients can also receive an organ from a living family member. But the pool of such kidneys has not been fully utilized, because the blood types of many spouses and siblings do not match those of their ill loved ones. Those willing kidney donors are lost to the system.
To keep those donors, Zenios and his colleagues at first tried to fashion a paired exchange system in which pairs of donors and patients would swap kidneys to make more matches. A spouse donor would give her kidney to an unrelated patient who matched her blood type. And that patient's mate would provide a kidney for the first donor's ill spouse. But when they ran the probabilities, Zenios, who is associate professor of operations, information, and technology at the Business School, and University of Chicago medical ethicist Lainie Ross, M.D., found that such a live donor exchange would only result in 1 percent more transplants. The hitch is that certain blood types are easier to match than others. For example, O-negative types are universal donors but not universal recipients. A pair with an O-type spousal donor and an A-type mate would rarely have to wait. But a pair in which the mate needs an O-type kidney will have to wait for a match. It is difficult to find suitable pairs. The conclusion: "It's more desirable to have an indirect exchange with the cadaveric waiting list-a procedure for which there is no mechanism at the moment," says Zenios.
Now, in their most recent analysis, Zenios and Ross, along with transplant surgeon E. Steve Woodle, M.D. at the University of Cincinatti, demonstrate how about 10 percent more kidneys-1,000 to 2,000-could be made available each year. The idea is that once a living donor gives a kidney to the pool, her loved one rises higher on the transplant list and receives the first kidney that matches. "It's as if they are buying priority," Zenios says. "We figured there would be a decrease in waiting time up to 15 percent, which also translates to an improvement in life expectancy of 15 percent. Some of those who normally have to wait so long will have a chance to receive a kidney before their health completely deteriorates," says Zenios. "We remove some of the gridlock from the current system."
Zenios and his colleagues have spent months painstakingly devising a statistically sound distribution mechanism that is fair to all patients. The system takes into account many complex factors-such as general health condition, age, imminence of total kidney failure, as well as blood type-that are used to rank waiting recipients.
But the research is not without controversy since it almost suggests a market for organs. "You're not buying organs, but you are in a sense purchasing a place on the list," says Zenios. "It has a bit of a market flavor, and physicians have a very high threshold for ethics due to their oath of primum non nocere, or 'first do no harm.'" In fact, a heated debate has swirled around organ distribution for nearly a decade, partly because it has generated inequalities for the African-American population, in which O-type blood is found in disproportionately high numbers compared to Caucasians.
In his modeling, Zenios checked to see if anyone would be any worse off due to the proposed changes. He found that there is a small possibility that O-type candidates who do not bring donors to the pool might be worse off. That's because O-types who bring a donor to the pool rise to the top of the O waiting list. Unlike other blood types that can match with multiple types, O-types can match only with other Os. The O-types who are unable to bring donors are stuck waiting behind the O-types who brought donors. They have fewer options and are likely to wait longer, unless the indirect exchanges involve sufficiently many O-types.
Although both O-positive and O-negative blood types are affected, O-negative types will be at a greater disadvantage than O-positive types.
The idea still generates many more kidneys overall. Zenios expects that medical experts will test his work over the next couple of years before any changes are attempted.
###
RSS