There exist a normal distribution of deaths from lung cancer for smokers and another for non-smokers; these life expectancy charts are the bread and butter of insurance companies; and a great deal of mathematics goes into them.
When there exists a disparity in these numbers, contributing factors are found. One of the primary candidates is SHS. Statistically speaking, the number of non-smokers who die of smoke related deaths is lower than that for smokers. (Duh!) The SHS group would be expected to be in the same number range as the Non-smoking group, however it isn't. The diffenence between deaths in the non-smoking group and the smoking group, for non-smokers is attributed to the only common link, SHS.
I'm with you on the normal distribution of deaths from lung cancer. But I still don't understand how they determine how many deaths were caused specifically by SHS as opposed to other environmental factors that could contribute to lung cancer.