Applying conversational artificial intelligence as a behavioral intervention to improve medication adherence
Abstract
Although two-thirds of Americans have a prescription medication, approximately 50% of people do not take their medications as prescribed. Of all medication-related hospital admissions in the United States, up to 70% are due to poor medication adherence, with a resultant cost of approximately $100 billion a year.
The patient-centered approach to medical consultations is critical to eliminating these addressable issues. In this collaborative approach, the patient and the health care provider share decision making, significantly motivating the patient to comply with the agreed upon health behavior. Tailored content is based on the individual’s beliefs, needs, and resources in regard to a given health outcome. Hereafter, strategic health messages can be created using evidence-based theories and delivered through a chosen channel, such as a virtual assistant.
In this project, the most appropriate content is identified and delivered, thereby creating a personalized, motivational, and reinforcing touchpoint.
Approach
The aim of this project is to create and test tailored interactions for automated conversations using a virtual assistant. This research focuses on developing behavioral interventions by exploring the relationship between the patient’s provider consultation, voice technology and the patient’s adherence to prescribed dose regimens.
Outcome
Barcenas & Brandin is actively supporting this ongoing research to validate whether tailored content increases the personal relevance of the messages and is perceived more positively. Compared to generic information, this way of communication has proven to be more effective in regard to influencing health behavior.
Research lead
Research partners
Industry partner
© Barcenas & Brandin LLC 2021
San Francisco
665 3rd Street
San Francisco, CA 94107
United States
Copenhagen
Frederiksberg Allé 25,
1820 Frederiksberg, Copenhagen
Denmark
Contact
hello@barcenas-brandin.com