I prefer to think that this happens when you pick a measure that ignores key aspects of the results. For example, would the results be the same if the mortality

I prefer to think that this happens when you pick a measure that ignores key aspects of the results. For example, would the results be the same if the mortality rate was complexity-weighted?