Here are some cool animations I made a couple years ago for teaching DAG-based causal inference

Confounding!

Reporting of cluster randomised crossover trials: extension of the CONSORT 2010 statement with explanation and elaboration This article presents the CONSORT (consolidated standards of reporting trials) extension for cluster randomised crossover trials. A cluster randomised crossover trial involves randomisation of groups ...

New CONSORT extension for cluster-crossover trials published in @bmj.com

www.bmj.com/content/388/...

This is way too pretty to look at. The paper is well outside my wheelhouse but I would just peer review it favourably on how nice this is.

Prepare Reproducible Example Code via the Clipboard Convenience wrapper that uses the rmarkdown package to render small snippets of code to target formats that include both code and output. The goal is to encourage the sharing of small, reproducible, a...

A tale as old as time:
- Oh no! Code is broken, something must be wrong 😰
- Create reprex 🧑‍💻
- Find solution in the process of creating reprex 🥳

Help others help you, and maybe help yourself in the process: reprex.tidyverse.org

Simulating Data From Marginal Structural Models for a Survival Time Outcome Marginal structural models (MSMs) are often used to estimate causal effects of treatments on survival time outcomes from observational data when time-dependent confounding may be present. They can be...

New paper from Shaun Seaman and me on how to simulate data from marginal structural models (MSMs) for survival outcomes, including Cox MSMs. This can be useful in simulation studies evaluating causal inference methods that use MSMs. R code provided. onlinelibrary.wiley.com/doi/10.1002/...

New lingo for old problems?

I discuss the implications of the estimands framework for old problems of working with aggregate level data using a well known intercurrent (post-randomization) event in treatment switching (cross-over) in oncology as an example

#Estimands
#EvidenceSynthesis

I would love to see the discussion of estimands and ICH E9(R1) for conducting meta-analysis. The language used in industry trials will be moving this way. I have seen very little recognition of this for those that regularly conduct aggregate level meta analyses