Measuring the efficiency of health services using Data Envelopment Analysis is a customary practice nowadays. In addition, some of the studies use analytical techniques to discover the variables that explain the efficiency level of such services. Nevertheless, efficiency analyses have seldom resulted in proposals to design organizational health care structures that provide adequate management for improving efficiency. Thus, to address the lack of use of Data Envelopment Analysis results in practice, this paper provides a general methodology based on the idea that measuring the efficiency of healthcare services must be complemented with their redesign to obtain potential efficiency improvements. This redesign includes healthcare processes or clinical flows, analytical techniques to advise such processes -e.g., OR/MS and AI models-, information technology for their support, and new management structures to facilitate the implementation and operation of the redesigns. Such methodology was applied to a large hospital specialized in cancer to illustrate its use and the results that can be achieved. They show that the potential cost savings of the proposed design are about US$ 1.0 million yearly. To generate such savings, the author designed new medical and management practices with the hospital personnel participation. The implementation of the practices and their continuous evaluation and innovation was possible by the new organizational structure designed for such purpose. This work shows the importance of integration methodologies to generate better-founded designs and facilitate their implementation. This integration is lacking in health-service design research, particularly in connection with architecture, processes, and organizational structure; this paper shows that this is a promising line of application. Highlights DEA analysis determines efficiency of health services and ways to improve it Scarce research on how to use DEA results to improve health processes exists A DEA-based methodology to design health and supporting flows is presented The methodology is applied to the design of the flows of a cancer hospital