The full text is available for free. http://www.thelancet.com/journals/lanres/article/PIIS2213-2600%2815%2900521-4/fulltext http://www.thelancet.com/journals/la

The full text is available for free. http://www.thelancet.com/journals/lanres/article/PIIS2213-2600%2815%2900521-4/fulltext http://www.thelancet.com/journals/lanres/article/PIIS2213-26... You need to log in, I used http://bugmenot.com/view/thelancet.com http://bugmenot.com/view/thelancet.com Having a quick skim, here's what I currently guess is going on. It's a meta-analysis, and most of the studies it looks at include smokers who aren't trying to quit. The effect remains (but weaker) when they restrict to studies of smokers who are interested in quitting. But "interested in quitting" doesn't necessarily mean they were actually trying to quit. So my guess is Simpson's paradox: for people who try to quit, e-cigs help them quit; for people not trying, e-cigs make approximately no difference; but people trying to quit are less likely to use e-cigs, and so people using e-cigs look like they're less likely to quit. Reading the paper more thoroughly might be able to falsify this guess. ("Quit" here means "quit cigarettes". Someone who stops smoking cigarettes, but continues smoking e-cigs, is considered to have quit.)