Is it cost effective to treat the world's poor?

Is it cost effective to treat the world's poor?

Is it cost effective to treat the world's poor?

Health and medicine explained.
July 17 2007 12:02 PM

Wrong Number

Is it cost effective to treat the world's poor?

Joseph Jeune before treatment. Click image to expand.
An HIV-infected boy in 2003 before drug treatments

Last month, I received a desperate e-mail from a former medical student of mine, now working in a remote Zambian medical clinic. A 3-year-old boy weighing only 15 pounds—about the same as a typical 6-month-old—was admitted to the never-empty malnutrition ward. The student found the boy's heart made a loud rushing noise instead of the usual lub-dub, indicating a serious malformation of the heart. Ruing the lack of surgical options, she asked, "[A]re there any inexpensive drugs we could give?" Sadly, I wrote back, there are none. Without heart surgery, the child almost certainly will die.

Pediatric heart surgery is fabulously expensive, often costing hundreds of thousands of dollars per case in the United States. Thus it would be foolhardy, goes the thinking, to offer surgery to poor African children who live on less than a few dollars per week. Isn't it better to invest in more cost-effective public-health measures, like mosquito netting to prevent malaria and vaccines against diarrhea? For decades, this kind of reasoning has been used to deny expensive but lifesaving treatments to the world's poor, most notably for HIV infection, in favor of more cost-effective measures focused on prevention. Dollar-for-dollar prevention is supposed to yield greater aggregate quality-of-life benefits than actual treatment.


Yet this seemingly reasonable argument is not only weak, but unfairly rigged, so the world's poor can never win.

The case against expensive treatments hinges on a statistic that economists use to compare medical interventions, called the "quality-adjusted life year." Here's how the sausage is made: How good a person feels is ranked on an arbitrary numeric scale and amortized over time. One QALY means the same thing as one year of perfect health, two years of half-perfect health, or four years of one-quarter perfect health. QALYs are an exchange unit that theoretically permits comparison between apples and oranges. Take a hypothetical example: pills that cost $200 for diabetics and give two years of half-normal health, versus a pancreas transplant that costs $5,000 and gives 10 years of one-fifth-normal health. The pills cost $200 per QALY, while the surgery costs $2,500 per QALY. If you believe that anyone really can tell half-normal health from one-fifth-normal health—and if you're a health economist, you probably do—the pills deserve funding more than the transplant.

Joseph Jeune after treatment. Click image to expand.
The same boy, holding his neice six months later, after drug treatments provided by Partners in Health

To ration care, a government or insurer determines how much a QALY is worth, and cuts health services with costs above where that line is drawn. This methodology hasn't really changed medical care in the United States because the threshold for QALY-based rationing is set high. In an unprecedented 1972 decision to fund a specific medical problem, Medicare began paying for kidney dialysis, which costs roughly $50,000 per QALY. In effect, this created a de facto cost-benefit threshold, and people have gamed the system ever since. It's not hard. As the British Medical Journal pointed out last year, most published studies of medical treatments in the United States find that all manner of medical treatments cost— voilà!—less than $50,000 per QALY. This also goes for HIV treatment in the United States. In 2001, for example, a group of Harvard researchers estimated in the New England Journal of Medicine that HIV medicines cost roughly $13,000 to $23,000 to give somebody a single QALY. The authors concluded the drug treatment for HIV was "highly cost effective and should be made available to all patients who can benefit from it."

What's easily affordable in rich countries, though, seems out-of-reach in poor ones where rationing thresholds are lower. Unfortunately, needy nations are stuck with figures conjured for countries where the ceiling is $50,000 per QALY. And the illusion of unaffordable treatment in poor areas is further bolstered by another insidious feature of QALY-based economics. In an experiment in the early 1990s, Oregon sought to ration health care by ranking all medical treatments, based on which yielded the most QALYs for the buck. Several odd findings emerged. Most notably, treatment of thumb-sucking and certain dental problems placed higher than treatment for cystic fibrosis and AIDS. Once these findings hit the media, Oregon abandoned the project and never rationed care.

What does thumb-sucking in Oregon have to do with AIDS in Africa? Oregon's experience showed how small improvements in huge numbers of relatively healthy young people (for example, millions of kids whose happiness is 1 percent higher since they don't thumb-suck, whatever that means) rack up QALYs faster than big improvements in a small number of really sick people (for example, a few dozen chronically ill folks with cystic fibrosis whose happiness is 10 percent higher from pricey antibiotics to treat their pneumonias). That's why—to economists, if to almost nobody else—it seemed fine to rank the treatment of thumb-sucking over cystic fibrosis, since it yielded the greatest overall QALY benefit.

With this logic, Africans with AIDS and expensive heart troubles never will qualify for life-saving drugs or surgeries. As the thumb-sucking example shows, QALYs dictate that paying for cheap preventive care will always win over expensive treatments, like those for AIDS. In 2002, the Lancet reviewed the medical literature and concluded that preventing HIV transmission to uninfected Africans (for example, via condoms distribution and education) was almost 1,000 times more cost-effective at generating QALYs than treating AIDS victims with expensive antiviral medications.

So, QALYs are unscientific, subject to powerful bias, recklessly applied out of context, and inherently biased toward prevention and away from treatment. But they make for devastating—and powerfully misleading—sound bites. The argument comes down to this: Doesn't it seem absurd to pay $23,000 to buy drugs to get one year of good health for an African with AIDS, when you can get the same amount of wellness for 1,000 Africans for the same cost?