Creativity broadens search for cures

"Some people believe you're either born creative or you're not," says Harvey Arbesman, M.D., M.S., clinical assistant professor in the department of dermatology and the department of social and preventive medicine, University of Buffalo School of Medicine.

"But people have studied and demonstrated over many years that the ability to increase your creative thinking capacity can be taught.

Disease example One example involves Lyme disease, which can be associated with skin manifestations.

Initially, Dr. Arbesman says, "We didn't have the hypothesis that they were (caused by) an infectious agent, so we didn't treat those skin manifestations with antibiotics. But now that we know they're related to Lyme disease, and Lyme disease has an infectious cause, we can treat them with an antibiotic. If you didn't have that hypothesis at first, you wouldn't be able to make a paradigm shift."

According to Stephen Hawking, a good hypothesis must describe a large class of observations based on a model that contains just a few arbitrary elements.

"Ideally, one wants to explain a lot of things with just a few ideas," Dr. Arbesman says. "A good hypothesis also must be able to make a definite prediction about the results of future observations. If it doesn't have predictive value - the idea that 'if we do this, we would expect that' - then it's not a good hypothesis."

A scientifically useful hypothesis also must be falsifiable, according to philosopher of science Karl Popper.

"If you say all swans are white and all you keep finding are white swans, that supports the hypothesis but doesn't prove it. But if you find one black swan, that disproves it," Dr. Arbesman explains.

Tools for generating hypotheses include the creative problem-solving process, which involves understanding the challenge, generating ideas and preparing for action (Isaksen SG et al. Creative Approaches to Problem Solving. Dubuque, Iowa; Kendall/ Hunt, 1994.). This process requires a dynamic balance between divergent thinking (generating options) and convergent thinking (selecting options that merit further exploration).

Clues in epidemiology Epidemiology also can provide sources of clues for hypothesis generation.

"There's not one factor that causes a disease, even one with an infectious cause. For example, the tubercle bacillus is not the only factor in tuberculosis; it is a necessary but not sufficient factor. There must be a susceptible immune system, and possibly social factors that depress the person's immune system. The idea of a causal web states that every disease is multifactorial," Dr. Arbesman says.

Similarly, no single factor is likely to cause a heart attack or melanoma. It's usually many factors interacting with one another.

"Sometimes when we look at it from an epidemiologic point of view," he says, "we can find clues that may give us a better understanding of the disease or perhaps ideas for therapy. You may be able to alter one of the branches on the web, so to speak, to modify outcomes."