Cardinality: Eric in the BRR working group made an interesting observation on the 8 April conference call: I think that consent should be 0..* not 0..1. Why should a subject only be have one agreement pertinent to a study - consent to treat and privacy consent seem a minimum in most cases. In our case, we add consent to release data to different groups of researchers. Because of limitations in our submitted data (we don't record consent to treat, for example), we can live with 0..1 but not everyone will want to constrain their submitters that way.