Health Sentinel: A media-scanning artificial intelligence deployed by the National Centre for Disease Control (NCDC) in 2022 has issued more than 5,000 real-time alerts on infectious disease events across India, according to a new preprint analysis (not yet peer-reviewed). The system, Health Sentinel, built by New Delhi–based WadhwaniAI, automates India’s Integrated Disease Surveillance Programme (IDSP), which traditionally relied on manual reviews of print, TV and online reports.
Since launch, Health Sentinel has processed over 300 million news articles in 13 languages, identifying about 95,000 unique health events nationwide. Public-health experts at NCDC shortlisted 3,500+ of those (˜4%) as potential outbreaks for verification and action. WadhwaniAI said that between April 2022 and April 2025, the platform generated 5,000+ real-time alerts to state and district authorities.
Also Read | India’s TB cases fall 21%, nearly twice the global decline
The study reports a 98% reduction in manual workload and a 150% increase in the number of published events captured since 2022 compared with prior human-only surveillance. In 2024, 96% of health events published by the national system were extracted by the AI; only 4% came from manual scanning. Epidemiologists remain “in the loop” to validate signals before escalation.
Health Sentinel targets a long-standing gap in event-based surveillance, rapidly sifting vast, multilingual media to spot unusual clusters. At the same time, India, as a signatory to the International Health Regulations (IHR), must maintain timely national surveillance. Faster detection aims to trigger earlier field investigation and control measures.
Separate Indian research underscores the value of augmenting traditional, provider-reported (“passive”) surveillance. A 2023 pilot in private hospitals in Kerala’s Kasaragod district used algorithms to analyse acute febrile illness records and verified 10 events (nine outbreaks, including dengue and COVID-19). Broader reviews have likewise found that online news and social media data, coupled with machine-learning methods, can improve early outbreak detection and fill reporting delays.
Study authors caution that Health Sentinel’s findings are preliminary pending peer review. Still, the results suggest that combining automated media analytics with expert verification can expand reach, accelerate reconnaissance and reduce costs, key advantages as India contends with frequent vector-borne and respiratory-pathogen surges.