Those decisions should be guided by "financial portfolio theory," argues a paper published Wednesday in the science journal PLoS ONE. Applying portfolio theory to how the National Institutes of Health (NIH) allocates its $30 billion annual budget could cut the total years of life lost by patients by anywhere from 28 percent to 89 percent, the researchers calculate.

"We need a framework for deciding how to make allocations for biomedical research that is transparent, objective, rational and reproducible," Andrew Lo of the MIT Sloan School's Laboratory for Financial Engineering said in an interview. Lo, considered a leading authority on financial theory, co-authored the paper with MIT's Dimitrios Bisias and surgeon James Watkins of Brigham and Women's Hospital in Boston.

"What we're hoping is that scientists and policy makers collaborate on measuring success" and use portfolio theory as a guide for how much to invest in research into various diseases, he said.

The paper's basic concept was dismissed by patient advocates contacted by Reuters. Some pointed out that under the criteria the new analysis proposed, research into chronic diseases, rare disorders and illnesses like Alzheimer's that have eluded a cure would be deemed nearly worthless. NIH officials were not available for comment.

Broadly speaking, portfolio theory provides guidance on how much to allocate to different investments - stocks, bonds, oil futures, real estate - based on their risks and expected rates of return, or reward.

Just as investing in U.S. Treasuries poses different risks and promises different rewards than investing in coffee futures, government investments to combat one disease versus another yield different results.

Ordinarily, when portfolio theory is used to optimize risk/reward trade-offs, the rewards are dollars. The new paper argues that in biomedical research the rewards that count are reductions in years of life lost to disease.

The most efficient allocation of biomedical research dollars is that which maximizes years of life saved per dollar spent, the researchers argue. This means that curing a 2-year-old's brain tumor is worth 70-some years of life, while curing an 85-year-old's arthritis is worth essentially zero.

The researchers are quick to caution that the emphasis on years of life saved is only the first step. Once the application of portfolio theory to biomedical spending decisions is refined, additional criteria -- such as whether a given line of research reduces pain and suffering -- would also be factored in.

Biomedical researchers and patient advocates contacted by Reuters expressed several concerns about the recommendations.

One is that the years-of-life-lost approach fails to capture chronic diseases that do not cut lives short but do cause immense suffering. "How quickly a disease kills, and how many it kills, is only one criterion," said Scott Johnson, president of the Myelin Repair Foundation, which supports research aimed at curing multiple sclerosis. "Suffering and disability also count."

Rare diseases would also lose out, since (by definition) they affect so few people. Using years of life lost "is outrageous and completely crazy," said Susan Weiner of the Children's Brain Tumor Foundation.

SHIFT IN FUNDING

Currently, the NIH allocates funding based on five criteria: "public needs"; the scientific quality of the proposed research; how likely a field is to make scientific progress (based on the foundation of existing knowledge and the number of qualified researchers); the need to have a diversified research program; and the need to support such infrastructure components as people, equipment, and facilities.

Using years of life lost as the measure of return on investment, some of NIH's investments have performed like Google's meteoric stock rise right after its initial public offering. Others have languished like Bank of America since 2007.

For instance, the study shows that spending on HIV research had a return of 9,100 percent, reflecting the fact that HIV/AIDS was originally near-universally fatal, and killed young people. Now protease-inhibiting drugs have turned it into a more manageable disease that doesn't cut as many lives short. "They really shot the lights out with protease inhibitors," said Watkins.

By the same analysis, cancer had a 50 percent return, while heart disease had a 1,000 percent return. Brain diseases, such as Parkinson's and Alzheimer's, had a negative rate of return: neither has been cured or even treated successfully despite the millions of dollars spent on research, and their incidence is rising.

Applying portfolio theory would drastically change how NIH spends the public's money. The Institute of Child Health and Human Development, whose research stands out for how many years of life it has saved, would get 23 percent of NIH's money rather than the current 7 percent. The Heart, Lung and Blood Institute, whose research has led to statins and other life savers, would go from 17 percent to 23 percent. The Cancer Institute would get 52 percent rather than today's 27 percent.

The losers would be the institutes that focus on diabetes (whose rate has soared, yielding a negative return on investment); the brain (a poor track record in terms of saving lives lost to Alzheimer's and other diseases), and mental health. All would be nearly zeroed out.

The authors wrote that they "would not suggest adopting a policy that would eliminate funding for any disease group." Even the riskiest investments sometimes pay off, which is why some investors buy pork bellies rather than, say, the S&P 500.

LIMITED APPLICATION

The researchers readily admit that years of life is a crude measure of success. Research on plastic surgery for burn victims, said Watkins, would lose out because it does not necessarily extend life, even though it greatly diminishes patients' suffering. Eventually, he said, NIH and policy makers would also take into account improvements in quality of life.

Another problem with applying portfolio analysis to NIH is that past performance is no guarantee of future returns. Researchers have made no progress against Alzheimer's, but maybe the big breakthrough is just around the corner.

Since NIH typically funds the most basic research it is not easy to predict returns. Authors of the new paper admit "the relation between NIH spending and changes in YLL is not nearly as direct as the relation between investing in the stock market and changes in one's net worth."

Also, portfolio theory assumes that if $1 invested today will yield $1.20 in five years, then $2 will yield $2.40. Doubling the investment doubles the return. But with NIH, the authors acknowledged, "it is not at all clear that doubling the budget of a given Institute will double its impact on future changes in YLL."

But even this rough analysis showed how much the NIH gets for its spending. "If policymakers truly understood the implications of our empirical results," said Lo, "they would be pushing for greater appropriations for the NIH expeditiously."

(Reporting By Sharon Begley and Jonathan Spicer; Editing by Michele Gershberg and Diane Craft)

• A Permanent Economic Emergency