This minimizes the significance of home fires caused by smoking, which suggests that home fires caused by smoking were not a significant source of home fire deaths. This could be why a decline in smoking would not necessarily lead to significantly fewer home fire deaths.

If these fires often break out after people in the house fall asleep, that increases the danger posed by these home fires. This makes the discrepancy harder to explain, because we’d expect a reduction in home fires caused by smoking to lead to a comparable decline in deaths.

This suggests that the general decline in smoking is not occurring among the kind of person who smokes in bed. This could be why there hasn’t been a corresponding decline in deaths from home fires.

More deaths from home fires that began outside of bed increased, which could explain why overall deaths from home fires didn’t go down as much as we’d expect from a reduction in home fires caused by smoking in bed.

This explains how, even if home fires from smoking went down, because deaths per home fire have gone up, overall deaths from home fires might not have declined as much as expected.