Questions about trial sequential analysis for non-randomized control studies (non-RCTs)


New Member
May I ask a question on the trial sequential analysis (TSA)? It is a method of cumulative meta-analysis.

Could we pooled adjusted risk ratios (aRR) in trial sequential analysis, especially for observational studies in which there are obvious differences between crude RR and adjusted RR?
Now, I use TSA Viewer (Version Beta) to do it.
Free software:
User manual for TSA:

Please look at the attached figure, it seems that only numbers of event and total subjects can be entered in it. Thus, the calculated RRs/ORs are crude measures without any adjustment. I am not quite sure whether this method/this software can pool the adjusted RRs/ORs, especially for observational studies.

Is anyone familiar with this? Better to provide reference paper to me if it is convenient for you.

Thank you very much!
Last edited: