Bacon sandwiches are big news right now in the UK. A staple of the English diet,recent studies demonstrating increased cancer risks associated with eating bacon sandwiches have dismayed the great British public.
Sadly, the quality of the reporting surrounding this research has been, at best, rather patchy. For relative risks offer rich prospects for the hack journalist.
A study suggests that a bacon sandwich a day increases your risk of bowel cancer by 20 percent. Scary stuff! (This is particularly worrying news for those living in my corner of the world. Here, in Northumberland, we indulge in the Northumbrian bacon buttie - a potent mix of bacon, jam (that's jelly to our transatlantic cousins) and bread.)
Time to drop the beloved bacon buttie from the menu?
Not just yet. For to evaluate this risk we need to know the background rate of bowel cancers. Of 100 people we might expect 5 to get bowel cancer anyway. So, a 20 percent increased risk means the five in 100 people we would expect to get bowel cancer anyway will go up to six. In other words just one in 100 will be harmed in this way. So maybe the odd bacon buttie isn’t going to kill us after all.
Whenever you see a relative risk in a newspaper headline you need to start looking for the background risks. If they’re not reported then the chances are they’re quite small. The use of relative risks -- 20 percent increase, 30 percent decrease and so on -- makes effects look bigger. It's an example of the "framing" of a risk. Unscrupulous journalists manipulate the frame to sensationalise health risks.
Always remember that even if you double a very small number it isn’t going to make it much bigger. Twice almost-nothing is still almost-nothing, Spiegelhalter’s Law of Small Risks!
Never have bacon sandwiches so exercised the British nation.
More on risk from David Spiegelhalter