ADELA: a conversational virtual assistant to prevent delirium in hospitalized older persons

Abstract

Delirium is a sudden mental state that causes confusion and disorientation, affecting a person's ability to think and remember clearly. Virtual assistants are a promising alternative for non-pharmacological interventions. This research aims to present a prototype of ADELA, a conversational assistant to prevent delirium in hospitalized older persons who speak Spanish. A co-creation process with medical experts to identify requirements was carried out to later develop the assistant iteratively and evaluate it from a technical and usability perspectives; the latest using the Spanish version of the System Usability Scale (SUS) and the Chatbot Usability Questionnaire (CUQ), supplemented by qualitative data. Mean values of 75.5 and 85.94 were obtained for SUS and CUQ, respectively. The technical evaluation helped defining the minimum environmental specifications required for deployment. Obtained results imply the assistant is usable and potentially accepted by the target population; useful information was extracted for refinement. Technical evaluation showed positive results, indicating it can be used in a real clinical environment.