Traditionally science is based on the idea that if you work out the causes of a problem, you know how to solve it. But this reductionist view won't work for complex problems. Professor Diane Finegood discusses why we need a more integrative approach to produce the kind of evidence we need.