For the first issue, I think it’s just a rounding issue. 0.3% + 0.1% does not add up to 0.5%, but above that 0.5% + 97.9% does not add up to 98.3%. Some kind of weird rounding thing.
For the second issue, I’m not sure why you mean ‘shouldn’t it be 0?’ 0.1% + 0.1% adds up to the prior step of 0.2%. What this is saying is that you were at 0.2% at the step of green duloxetine, and then a portion (0.1%) had nothing else, and the other 0.1% went on to do something else (not end) but possibly it was ‘trimmed out’ and you don’t see it in the plot (we do a little trimming on the output for display and performance reasons). I don’t think you’ll ever see a zero value unless you have something like 50% of people from step 1 went on to something in step 2, and those same 50% people went on to step 3…the diff there would be 0%. But, we’ll never show a slice in the plot that represents 0% of people…we don’t draw something that has a zero value.