Comparing computer-interpretable guideline models: a case-study approach