Person-year estimation with incomplete follow-up
Project team: Heiko Becher, Volker Winkler
Funding: own funds
A standard parameter to compare the disease incidence of a cohort relative to the population is the standardized incidence ratio. For statistical inference is commonly assumed that the denominator, the expected number of cases, is fixed. Incident cases can sometimes be obtained by linkage with a disease registry, and a complete follow-up may not be possible. In that case, end-of-follow-up date and therefore, exact person-years of observation may be unknown.
We have developed a method to estimate the observation times and to derive the expected number of cases. Current results show that the additional variance induced is small, so that standard methods for inference can be applied. We investigate the impact of the underlying assumptions with a sensitivity analysis and illustrate the method with a numerical example and with a recent study.
We applied the method in the estimation of the standardized cancer incidence ratio in a cohort of migrants relative to the German population.