You could be reading it wrong. Note that it explicitly says "life expectancy at birth". There has been a lot of criticism of this sort of things from statistics-enabled researchers, who point out that almost all of the life-expectancy gains in the past century have been through elimination of most early-childhood deaths. Life expentancy at birth has increased, but the life expectancy of someone 30 or 60 years old hasn't actually changed much.
There has been a bit of publicity around related topics lately. Thus, there has been a lot of discussion of the apparent fact that the increase in mammograms has produced no measurable increase in lifetime, just an increase in medical bills for the testing (and the "treatment" of false positives ;-). Similar statistical problems have been reported for prostate-cancer screening, and for an assortment of other medical tests.
Another statistical trick used to make things look better than they are is the common practice of giving cancer survival rates in terms of survival 5 years after diagnosis. This means, for example, that if you were to come up with a new test that diagnoses a cancer 5 years earlier than any existing test, your test would result in a 100% "cure" rate even with no further treatment, and no change in the death statistics. I've heard a couple of interviews in which the interviewer points out this problem, and the interviewee just continues talking about the same "5-year survival" figures.
In general, it seems that if you're over 10 years old, modern medicine really hasn't done much in increase your (statistical) lifespan, though it is sometimes fairly good at extracting money for treatments that don't increase lifespan.
(Perhaps some of the treatments improve quality of life, but the statistics for that don't seem to be widely studied or reported. It might be interesting to be shown wrong in this regard, however. OTOH, there has been a bit of media coverage lately of the problems with "treatment" of false positives.)
(And a more general problem here is that the general public -- and the media -- is generally ignorant of even the most basic statistical concepts.)