Follow-up, I just got done comparing two groups' infection rates. So I actually conducted what would be risk differences between the groups. So in frequentists approach you would calculate risks per group and find difference and calculate the 95% CI. what I did was provide a flat prior for the differences, saying the two groups don't differ, and simulated the distribution of both groups' infection rates 10,000 times. I then subtracted the two rates for the 10,000 samples and ended up with a 95% credible interval of the differences. So I had 10,000 differences plotted as a histogram and I could see Group A had a >99.9% probability of having a worse infection rate than Group B, since 99.9 of their risk values were greater than Group B's risk values. This is probably the simplest application of bayes in biostats. Side note, patients weren't randomized to groups, so for a more conclusive analyses I would run this same thing in a model, controlling for patient characteristic impacting the outcome.