Hello,

I am relatively new to STATA and I am working on a research study investigating medical outcomes (binary dependent variables) with and without administration of a medical drug (binary independent variable). I would like to perform an instrumental variable analysis using a binary instrument (physicians who use the drug in 0-30% of their surgical cases [0] vs. physicians who use the drug in 90-100% of their surgical cases [1]). I have read through some posts on this forum and there are conflicting opinions on which command to use since STATA does not have an ivlogistic feature. Our study reports our results using logistic regressions. Which command would be most appropriate to perform the aforementioned task? I have seen some recommend ivregress and other recommend ivprobit. My understanding is that the use of ivregress is appropriate for a continuous outcome of interest.

Any guidance would be greatly appreciate, thank you so much in advance.

Amit