Mode transition in any inappropriate mode can be a common cause of any mishap in a complex health-care system. This paper presents an approach for formalizing and reasoning about optimal mode transition in a health-care system that uses several operating modes in various operating states. Modes are formal- ized and their relation to a state-based formalism is established through a re- finement approach. The efficiency of this approach is presented by formalizing an ideal operating mode transition of a cardiac pacemaker case study. An incre- mental approach is used to develop the system and its detailed design is verified through a series of refinements. The consequence of this approach is to improve system structuring, elicitation of system assumptions and expected functionality, as well as requirement traceability using modes in state-based modeling. Models are expressed in EVENT B modeling language and validated by a model checker tool: ProB.