Spoken dialogue systems have been increasingly employed to provide ubiquitous automated access via telephone to information and services for the non-Internet-connected public. In the health care context, dialogue systems have been successfully applied. Nevertheless, speech-based technology is not easy to implement because it requires a considerable development investment. The advent of VoiceXML for voice applications contributed to reduce the proliferation of incompatible dialogue interpreters, but introduced new complexity. As a response to these issues, we designed an architecture for dialogue representation and interpretation, AdaRTE, which allows developers to layout dialogue interactions through a high level formalism that offers both declarative and procedural features. AdaRTE aim is to provide a ground for deploying complex and adaptable dialogues whilst allows the experimentation and incremental adoption of innovative speech technologies. It provides the dynamic behavior of Augmented Transition Networks and enables the generation of different backends formats such as VoiceXML. It is especially targeted to the health care context, where a framework for easy dialogue deployment could reduce the barrier for a more widespread adoption of dialogue systems.