CataractBot: An LLM-Powered Expert-in-the-Loop Chatbot for Cataract Patients
Pragnya Ramjee, Bhuvan Sachdeva, Satvik Golechha, Shreyas Kulkarni, Geeta Fulari, Kaushik Murali, Mohit Jain
TL;DR
CataractBot tackles the problem of unreliable or inaccessible health information by combining an LLM with an expert-in-the-loop verification pipeline and a hospital-curated knowledge base, delivered via WhatsApp in five languages with multimodal input. By generating instant, general answers and asynchronously verifiable expert edits, it builds trust and reduces the information burden on busy clinicians while maintaining privacy. The study deployed CataractBot at Sankara Eye Hospital in India with 49 information seekers and 6 experts, conducting mixed-method analyses that show high trust in doctor-verified responses, substantial use by attendants, and improved information access across literacy levels. The findings illuminate design implications for scalable, expert-mediated AI chatbots in healthcare, including workflow integration, workload management for clinicians, language support, and KB maintenance, with potential applicability beyond cataract care.
Abstract
The healthcare landscape is evolving, with patients seeking reliable information about their health conditions and available treatment options. Despite the abundance of information sources, the digital age overwhelms individuals with excess, often inaccurate information. Patients primarily trust medical professionals, highlighting the need for expert-endorsed health information. However, increased patient loads on experts has led to reduced communication time, impacting information sharing. To address this gap, we developed CataractBot. CataractBot answers cataract surgery related questions instantly using an LLM to query a curated knowledge base, and provides expert-verified responses asynchronously. It has multimodal and multilingual capabilities. In an in-the-wild deployment study with 49 patients and attendants, 4 doctors, and 2 patient coordinators, CataractBot demonstrated potential, providing anytime accessibility, saving time, accommodating diverse literacy levels, alleviating power differences, and adding a privacy layer between patients and doctors. Users reported that their trust in the system was established through expert verification. Broadly, our results could inform future work on expert-mediated LLM bots.