Understanding bias when estimating life expectancy from age at death: A simulation approach applied to Morquio Syndrome A

Yin, X.; Ahn, J.; Boca, S. M.

Background: Life expectancy can be estimated accurately from a cohort of individuals born in the same year and followed from birth to death. Due to the difficult and time-consuming nature of following a cohort prospectively, life expectancy is often assessed based on death data, which may lead to potentially biased estimates. This is more likely to be a problem in rare diseases such as Morquio syndrome A. Method: To investigate how accurate the estimation of life expectancy is using death data, we simulate the survival of individuals with Morquio syndrome A under four different survival scenarios. In each scenario, we estimate the mean and median survival times within a defined period and compare them with the true life expectancy. Results: When life expectancy is constant during the entire period, using death data does not result in a biased estimate of life expectancy. However, when life expectancy increases during the follow-up period, using only death data leads to a substantial underestimation of life expectancy. Conclusion: Life expectancy can change over time, along with changes in the environment and/or biomedical innovation. When the life expectancy is increasing --- as is often expected to be the case in rare diseases --- estimating it based on contemporary death data will result in a downward bias. Therefore, it is crucial to understand how estimates of life expectancy are obtained and to interpret them in an appropriate context, and to assess estimation methods within a sensitivity analysis framework, similar to the simulations performed herein.