Reply to the Editor. Methodologic issues regarding background mortality in observational studies.