With a Little Help From My Conversational Agent: Towards a Voice Assistant for Improved Patient Compliance and Medication Therapy Safety

Jan Schulte to Brinke, Christian Fitte, Eduard Anton, Pascal Meier, Frank Teuteberg

In: Proceedings of the 14th International Joint Conference on Biomedical Engineering Systems and Technologies. International Joint Conference on Biomedical Engineering Systems and Technologies (BIOSTEC-2021) February 11-13 Vienna Austria SciTePress 2021.


The chronically ill and the elderly often need to take several drugs, which increases the complexity of medication management. This frequently results in a decrease in patient compliance and raises the risks of their drug therapy. To support patients in medication management, we developed a multimodal assistant that includes a conversational agent supplied with data from a database managed by healthcare professionals via a web service. The developed artifact analyzes medication plans, identifies adverse drug reactions and side effects, and reminds patients to take their medication correctly and on time. Applying the design science research paradigm, we systematically identified 16 issues, derived eight meta-requirements, and elaborated three design principles. Based on this, the artifact was implemented and evaluated by three experienced pharmacists, who highlighted the usefulness of the solution and provided feedback for further improvements. Finally, we present an evaluation concept for potential users and discuss the implications of the medication assistant. Overall, the medical assistant comprises valuable functionalities to support patients, and it increases medication therapy safety and patients’ compliance.

Deutsches Forschungszentrum für Künstliche Intelligenz
German Research Center for Artificial Intelligence