P21: Development and Experience with a Global Medical Information Vaccine Stability Chatbot
Poster Presenter
Sagar Shah
Global Medical Information Manager
Sanofi United States
Objectives
To discuss the development, deployment, and post-launch experience with a Global Medical Information (GMI) chatbot that provides accurate, scientifically validated information on vaccine temperature excursions intended for health care providers (HCPs) in the United States (US) and Germany.
Method
Sanofi built the chatbot in 2020 by creating a conversation flow to mimic live calls regarding vaccine temperature excursions. It was developed on a conversational artificial intelligence platform emphasizing agile ways of working. Post launch, conversations are monitored for medical accuracy.
Results
GMI successfully launched a free-text chatbot in Q4 of 2020 in both the US and Germany. This new channel is available to HCPs on Sanofi’s self-service medical information website 24/7. In the first 3 months after deployment, nearly 7% of visitors to the self-service website used the chatbot. Of the chatbot users, 24% received a vaccine temperature suitability answer from the bot in less than 3 minutes. The rest of the customers, who engaged with the bot had a non-covered inquiry type. It was also shown that 35% of users were accessing the chatbot outside of normal business hours. Daily review of the medical responses provided by the chatbot when used within scope, showed that 99% align with the MI response the provider would have received over the phone from the GMI call center. In addition, satisfaction ratings provided by the user throughout the conversation are an average of 4.3 out of 5.
Conclusion
GMI is responsible for responding to a large volume of inquiries annually. Nearly 10% of all inquiries received involve vaccine temperature excursions. Currently the most common channel HCPs utilize to address their temperature excursion inquiries is by calling GMI. However, as digital engagement grows in the medical world, providers are increasingly seeking digital experiences to address their needs. The chatbot program goals were to leverage a digital channel for HCPs to interface with that reduces the number of simple, recurrent inquiries reaching the GMI call center and has the potential to grow in scope over time. This project demonstrates that it is possible to automate an accurate, validated medical information response in an efficient and time saving manner using a chatbot. Customer satisfaction and feedback show that this channel is a suitable method to deliver medical information content that is accessible beyond normal business hours. The high satisfaction rating validates the customer centric process we used to build the chatbot and create the conversation flows.
