India needs national common health dataset to create AI models: VK Paul

India’s health secretary VK Paul urged the creation of a national common health dataset to support AI-driven public health initiatives, addressing fragmented and siloed data that hinders predictive modeling for outbreak risks, precision medication, and disease surveillance. Speaking at the Confederation of Indian Industry’s (CII) Annual Business Summit in New Delhi, Paul stressed that such a dataset would enable AI models only if validated at technical, clinical, and health levels. The focus must first be on strengthening India’s primary healthcare system to tackle non-communicable diseases (NCDs), which account for 63% of all deaths, according to Union Health Ministry data. Paul highlighted the need to leverage over 185,000 Ayushman Arogya Mandirs (AAMs) under the Pradhan Mantri Jan Arogya Yojana (PMJAY) for early disease detection and management through lifestyle interventions and therapies. Secondary prevention of NCDs—including cardiovascular diseases, hypertension, diabetes, fatty liver issues, and mental health conditions—must be prioritized through primary healthcare systems. Paul also called for expanding family welfare as a medical specialty to address the growing burden of chronic diseases, noting that India’s 5,000 MD-level family medicine doctors (0.36 per 100,000 people) are insufficient. He criticized the current medical training model, which fails to equip doctors with the continuity of care, second-line treatment, and monitoring skills needed for chronic disease management within a five-and-a-half-year MBBS program. Paul advocated for reimagining medical education to produce future doctors capable of treating patients across all age groups, currently scattered across pediatrics, internal medicine, and community medicine.