Natural Language Processing-Driven Insights from Social Media: Topic Modeling and Sentiment Analysis of Healthcare Sustainability Discourse

The transition to environmentally sustainable healthcare is gaining urgency, yet public discourse shaping this shift remains underexamined. This study employs natural language processing (NLP) to analyze 15,976 English-language tweets (2006–2024) related to sustainable healthcare. Using Latent Dirichlet Allocation (LDA), eight dominant topics were identified, including eco-friendly access, net-zero implementation, climate impact, emissions, cost and waste, education, infrastructure, and green technologies. Sentiment analysis (VADER) of 9433 tweets showed 59.1% positive, 31.1% neutral, and 9.8% negative sentiment, with AI and technology topics receiving the highest positivity (73.5%) and climate-related topics the most negativity. Thematic analysis of 800 tweets revealed six cross-cutting themes, including healthcare’s environmental responsibility, co-benefits for health, urgency of climate action, and optimism in technological solutions. These findings offer a nuanced understanding of public perceptions, informing targeted strategies and communication for healthcare sustainability. The study also demonstrates the value of mixed-method NLP in examining enablers and barriers to health system transformation.