We identified a range of existing PI that are valid for internal quality assessment as they are evidence-based, accepted by professionals, and reliable. For external use, problems of patient case-mix and sample size per PI should be better addressed. Further research is needed for selecting indicators that predict clinical outcomes.