Background: Survival trend analyses examine mortality outcomes over time. The impact of conducting survival trend analyses without accounting for improved population survival has not been systematically studied.
Methods: The 1-year risk of death in the 100 most common hospital admissions for Ontario adults in 1994, 1999, 2004, and 2009 was determined. Generalized linear models were used to determine if adjusted death risk changed significantly over time with and without accounting for population survival.
Results: The statistical significance of temporal trends in survival changed after accounting for population life expectancy in 16 diagnoses (16 percent) (in 13 of 55 diagnoses, statistically significant decreasing mortality trends became insignificant; in 3 of 15 diagnoses, insignificant trends changed to a significant increase in mortality risk over time).
Conclusions: These results highlight the importance of accounting for population life-expectancy changes in survival trend analyses.
Keywords: Relative survival; generalized linear model; survival trend analyses.
© Health Research and Educational Trust.