Comparative Outcomes with an Artificial Intelligence-Powered Short Message Service Chatbot After Total Joint Arthroplasty.

Abstract

Introduction: Access to a patient's surgeon has become increasingly difficult given increased volumes of total joint arthroplasty (TJA). A surgeon-specific conversational chatbot has the theoretical advantage of engaging and guiding patients through the perioperative journey irrespective of surgeon availability. The purpose of this study was to retrospectively assess if enrollment in a perioperative chatbot was associated with improved outcomes following TJA compared to a historical cohort not enrolled.

Materials and methods: We identified 1,338 TJA patients who enrolled in a Short Message Service (SMS) chatbot from 2020-2022 with greater than three months of follow up. Similarly, we identified a historical cohort of 1703 patients not enrolled in the SMS chatbot. The associations of chatbot enrollment on emergency department visits and readmissions within 30 and 90 days postoperatively were analyzed.

Results: Chatbot enrollment was associated with a significantly decreased odds ratios of emergency department visits at 30 and 90 days postoperatively (0.40 and 0.45 respectively, p<0.0001) compared to a historical cohort not enrolled. There was also a significantly decreased odds ratio of readmission by 90 days postoperatively (0.60, p=0.0168) for those enrolled in the SMS chatbot.

Conclusions: Chatbot enrollment was associated with decreased odds ratios of emergency department visits and readmissions following TJA compared to a historical cohort without access to a perioperative chatbot. Perioperative chatbots may provide an opportunity to further improve patient outcomes following TJA.