Core Insights - The article discusses the increasing pressure on global healthcare systems due to aging populations and chronic disease burdens, highlighting inefficiencies in patient care processes, particularly in China [2] - Recent studies demonstrate that AI chatbots based on large language models (LLMs) can significantly enhance healthcare delivery by improving patient engagement and streamlining care transitions [3] Study 1: PreA Chatbot - The PreA chatbot, developed to assist patients transitioning from primary to specialist care, was tested in a randomized controlled trial involving 2,069 patients across 24 departments in two major hospitals in Western China [6][7] - Key findings include a 28.7% reduction in consultation time for patients using PreA, with average times decreasing from 4.41 minutes to 3.14 minutes, and a 113.1% improvement in the perceived usefulness of referral reports by specialists [7] - The success of PreA is attributed to its co-design with local stakeholders, ensuring it meets real clinical needs and operates effectively in resource-limited settings [10] Study 2: P&P Care Chatbot - The P&P Care chatbot focuses on enhancing the primary care experience and was also tested in a randomized controlled trial involving 2,113 participants across 11 provinces in China [12][13] - The co-design approach involved community members, leading to features that cater to cultural and literacy needs, such as voice interfaces and offline capabilities [15] - The P&P Care chatbot outperformed traditional primary care methods in areas like history-taking, diagnostic accuracy, and chronic disease management [15] Common Insights - Both studies emphasize the importance of co-design in deploying AI in healthcare, which helps avoid systemic biases and ensures tools are aligned with actual needs [17] - AI tools like PreA and P&P Care are not intended to replace doctors but to handle routine tasks, allowing healthcare professionals to focus on complex decision-making and patient care [18] - The robust performance of these AI chatbots in resource-limited environments suggests they could serve as models for improving healthcare equity globally [19]