Cooperative clinical information systems (CCIS) are destined to become an invaluable tool for managing care processes in which a number of interrelated care tasks are performed by a network of cooperating agents. After introducing this important new domain, the key mechanisms and structures needed to support CCIS are detailed - a set of communication primitives are developed, the concept of agent commitment is described, the structures needed for reasoning about allocating tasks to appropriate agents are presented, and the mechanisms for managing change in the cooperation community are outlined.