Nicki,
The estimate was developed using data from the town of Marshfield, Wisconsin. Participation in the study was voluntary, and detailed data about illness, vaccination, and lifestyle was collected from those enrolled.
Because of this, my answer to your question is yes, those items were factored in, at least as far as the sample population is concerned. However, I see two additional issues with the estimate. The first is that the participants were self-selected so the study was not truly random. The second is whether or not the town is sufficiently representative of the rest of the U.S. to allow generalization of the 44% rate.
I suspect that the success rate of the vaccine was actually lower since 2 out of 3 strains in the vaccine were not the "right" ones (i.e., the strains that actually spread during flu season). Scientists have to guess each year and apparently this year they guessed wrong. I remember the "swine flu"
non-epidemic of 1975-76.... of course I'm showing my age there.