Agenda:

15.00–15.30 Welcome and Food Service

15.30–16.15 Heterogeneity and uncertainty in estimates of effect, by Lars Christian Stene

“Everybody” talks about heterogeneity, but there are so many different kinds of heterogeneity. “Everybody” talks about effects, but there are so many different measures of effect. I say most measures of effect have a lot of uncertainty, even in randomized trials. I will illustrate how scenarios with extreme variation in heterogeneity (and deterministic vs probabilistic scenarios) can give results that appear very similar, and how wide the 95% confidence intervals may be around measures of efficacy (effect) in real trials. Examples from the TN10 Teplizumab T1D prevention trial (K. Herold, NEJM 2019, approved for use in the USA, but points made are universal – I think).

16.15–16.30 Questions/Discussion

This is a hybrid event that allows for both in-person and online participation. A link has been distributed to everyone affiliated with Oslo Dibetes Research Center. If you haven’t receved an invite, please dont hesitate to contact us.