In the case of CMS56 where we need to count the number of lower body fractures, in theory, a clinician could document the same encounter diagnosis and problem list diagnosis on the same encounter, it seems to us that the measure intent is to count them only once. But there is no mechanism in CQL logic to handle diagnosis data elements from multiple sources that might ultimately be the same diagnosis.

Can you give us some guidance on how we should deduplicate diagnoses from different sources when it comes to count?