Large Language Models Used for Public Health Surveillance
Researchers are increasingly deploying large language models (LLMs) for public health "infoveillance" to monitor population health trends and misinformation. A paper in *npj Digital Medicine* details how a suite of LLMs can analyze digital conversations to identify emerging health needs and patient sentiment. This technique allows for real-time insights into public health concerns.
- Venture capital funding for U.S. digital health startups saw a 35% increase in 2025, reaching $14.2 billion. AI-enabled companies are attracting a significant portion of this investment, securing 54% of the total funding in 2025, up from 37% the previous year. - For consumer health apps, HIPAA compliance is not always mandatory and depends on whether the app handles Protected Health Information (PHI) on behalf of a "covered entity" like a healthcare provider. Many direct-to-consumer wellness apps fall outside of HIPAA's scope, making it crucial for founders to understand their specific legal obligations. - Successful consumer health apps often employ a multi-pronged user acquisition strategy that includes App Store Optimization (ASO), paid advertising on social media platforms, and content marketing to establish expertise. Building a strong community and encouraging user-generated content are also key tactics for driving organic growth. - Integrating with wearables is becoming essential for consumer health apps; however, each wearable manufacturer has its own API with different authentication processes and data structures. Unified API platforms like Terra, Validic, and Vitalera are emerging to simplify the integration of data from popular devices like those from Apple, Fitbit, and Garmin. - The longevity and biohacking sector is attracting significant investment, with a focus on areas like cellular reprogramming and AI-driven drug discovery. Notable startups include Altos Labs, which has reportedly raised $3 billion to reverse age-related decline, and NewLimit, which recently secured a $130 million funding round. - Analyzing patient sentiment on social media and health forums can provide valuable insights into user needs and frustrations with existing healthcare solutions. This data can inform product development and help create a more user-centric health app. - While LLMs show promise for public health surveillance, they also present challenges such as the potential for biases in the training data, a lack of transparency in their decision-making processes, and concerns around patient data privacy. Regulatory frameworks for the use of LLMs in healthcare are still in the early stages of development. - The U.S. mHealth app market was valued at $16.51 billion in 2024 and is projected to grow at a compound annual growth rate of 12.4% through 2034. The monitoring services segment, which includes apps for chronic disease management, held the largest market share in 2026.
