Reply to comment

In Health, Evidence Matters But So Does Behavior

January 18, 2007

Researchers are pretty unanimous today that the latest statistics on deaths caused by cancer show that a two-year downturn is no fluke and may represent a real trend.

Check the stats here, but remember that the state of science even in the 21st Century does not allow for making simple conclusions. Is the decline in cancer deaths due to better drugs or to behavior change that has made smoking passe and keeping appointments for mammograms and colonoscopies de rigeur?

There is a strong suggestion that it is, in fact, behavior that has made most of the difference. But we still don't know from these new statistics whether fewer cancer deaths signify medical breakthroughs or simply more people surviving – and suffering the chronic effects of – this most feared disease.

The key to this and every other medical question lies in the long and arduous process of amassing solid evidence. For an explanation of where we would be without the use of rigorous evidence collection, see Johns Hopkins professor Kay Dickersin's argument that none of the other 15 greatest medical milestones the journal BMJ listed today would have been possible without the notion that evidence matters.