Starting point of each emulation was a clinically relevant question formulated under the PICO framework and identified to equal parts through one of two possible processes. Our advisory board consisting of pharmacological experts compiled a list of currently relevant questions for which we tried to identify and recruit suitable registers. 
In addition, we contacted potentially suitable German registers and recruited those that were interested and offered to provide suitable data. Questions were then formulated by clinical experts chosen by cooperating registers. 

The basic PICOS are translated to more detailed target trials through iterative discussions with the clinical experts. If an emulation is feasible in principle using the available data, we create a study protocol and conduct a prospective risk of bias assessment. 

While we conducteach non-randomized registry-based cohort study, we record all feasibility problems and relevant methodological aspects. 

This is followed by a final bias assessment and a systematic comparison of randomized controlled trials (RCTs) on the same clinical question (if available) and our results in order to assess causes of deviation between the two.

Besides gaining insights on the feasibility of conducting such non-randomized studies using currently available German registry data we also aim to support the dissemination of state-of-the art statistical methodology. 
We develop materials like templates for all important steps of the analysis process and reusable R analysis scripts that will be made available to the public.