Healthcare systems globally are confronted with escalating demands. Physician assistants, critical in maintaining continuity of care for patients with chronic conditions, often struggle with high workloads, limiting their capacity to offer personalised, ongoing support. While AI holds promise in easing these pressures by supporting administrative tasks and decision-making, its integration also raises concerns. Key issues include AI’s potential to devalue the role of healthcare workers and its limited ability to deliver truly personalised and empathetic care—critical aspects that require further investigation.

To address these challenges, the project will evaluate a generative AI tool developed in collaboration with a Chinese telemedicine company. Customized for the healthcare domain using a leading large language model (LLM), the AI tool will assist physician assistants by generating suggested responses to medical queries from patients with chronic conditions. In the treatment group, assistants will have access to the AI tool, while those in the control group will continue with their current workflow. The study will examine the effect of AI on healthcare workers’ productivity and job satisfaction, as well as on patient-related metrics such as interaction quality, medication adherence, and overall satisfaction.

The findings from this study will offer critical insights into whether and how AI tools can enhance continuity of care while addressing the needs of both healthcare providers and patients. By measuring the causal effects on healthcare worker and patient outcomes and highlighting the potential benefits and risks, this research could provide actionable recommendations for healthcare organisations seeking to incorporate AI into their care delivery models.