Chennai, July 7, 2025 — In a milestone move for public health innovation, Tamil Nadu has emerged as the first state in India to officially deploy an artificial intelligence-based model that predicts the risk of death among tuberculosis (TB) patients. The pioneering effort aims to help healthcare providers identify and prioritize patients with severe forms of TB who require immediate medical attention, potentially saving countless lives.
The predictive tool, developed by the Indian Council of Medical Research–National Institute of Epidemiology (ICMR-NIE), uses clinical and physiological data collected during diagnosis to estimate the patient’s risk level. By integrating this AI-powered assessment directly into the state’s existing TB care infrastructure, Tamil Nadu is reinforcing its position as a national leader in healthcare innovation and digital intervention.
At the heart of this initiative is the TB SeWA (Severe TB Web Application), a system already used in Tamil Nadu’s flagship program for severe TB cases—Tamil Nadu Kasanoi Erappila Thittam (TN-KET). The AI model works by analyzing five key indicators: body mass index (BMI), respiratory rate, oxygen saturation, the presence of pedal edema, and the patient’s functional status at the time of registration. The algorithm then flags high-risk individuals and advises early hospital admission.
Dr. Asha Frederick, State TB Officer of Tamil Nadu, emphasized that this development is not just a technological upgrade, but a targeted public health tool that enhances clinical decision-making. “The AI model gives us a scientific way to quickly identify patients at high risk of mortality. It allows frontline workers to act faster and save lives by ensuring timely referrals and hospital care,” she explained.
What makes the Tamil Nadu model significant is its integration into field operations. As soon as a TB patient is diagnosed and entered into the state’s system, the AI model automatically processes the input data and alerts medical teams if the patient falls into the high-risk category. This early warning system, tested through pilot studies, has already demonstrated a tangible impact. In pilot districts, the model helped frontline workers reduce delays in hospital admissions and improve follow-up care coordination.
India remains the country with the highest TB burden in the world, accounting for nearly 28% of global cases. According to the World Health Organization, India recorded an estimated 2.8 million TB cases in 2023, with more than 400,000 deaths. Tamil Nadu, like many Indian states, has worked hard to combat TB through active screening and treatment campaigns, but challenges such as delayed diagnosis, malnutrition, co-morbidities like diabetes, and treatment default continue to affect patient outcomes.
Under TN-KET, launched in 2022, Tamil Nadu already identifies severely ill TB patients at the time of diagnosis and prioritizes them for nutritional, clinical, and psychological support. The new AI model adds another layer to this intervention, making the process faster, sharper, and more accurate. Health officials say it’s a significant step toward meeting India’s national TB elimination goal of 2025.
Medical professionals on the ground are already noticing a difference. In Chennai’s urban health centers and rural primary health clinics across districts like Tiruvallur and Madurai, health workers report that the AI-generated risk assessments are improving their ability to triage patients effectively. One health worker described how a flagged alert prompted an emergency hospital referral for a patient who otherwise seemed stable on paper. “We wouldn’t have caught it so early without the AI prompt,” she said.
The AI model doesn’t just support diagnosis—it also optimizes human resources. In busy clinics where doctors and field staff handle large caseloads, having a predictive tool to identify red flags ensures that time and attention are focused where they are needed most. According to Dr. Senthil Kumar, a pulmonologist at the Government Stanley Medical College, “It allows us to allocate resources better. We know whom to prioritize for hospitalization, for nutritional support, for close monitoring.”
Public health experts have lauded Tamil Nadu for leading this initiative, not only for being the first to adopt such a model but for doing so through public systems rather than pilot projects in academic or private settings. The fact that the AI model is already integrated into Nikshay—the national TB reporting platform—means it could be replicated in other states with relative ease, once training and digital infrastructure are in place.
Other states such as Maharashtra, Uttar Pradesh, and Odisha are watching closely. Some have expressed interest in piloting similar models, although none have yet achieved full-scale deployment like Tamil Nadu. Odisha, for instance, has launched an AI policy for broader healthcare applications, but it has not yet introduced this specific TB mortality risk model.
Still, challenges remain. Data quality and completeness are critical for accurate predictions. Health workers must be trained to input the required indicators accurately and consistently. Internet connectivity in remote areas may slow down real-time data processing, and there is a need to guard against over-reliance on algorithms when clinical judgment is still essential. But officials remain optimistic. “The model is a decision support system—it doesn’t replace doctors, but it strengthens their ability to respond,” said Dr. Asha Frederick.
Ethical concerns around patient data have also been addressed. The state confirms that all data used by the AI system is anonymized, and that the TB SeWA platform is protected by government-mandated security protocols. Patients are made aware of digital tracking and consent procedures as part of their registration.
This landmark effort is not just about saving lives—it’s about changing how public health systems think. Instead of reactive care, where doctors wait for complications to arise, Tamil Nadu’s system enables predictive care. The goal is to intervene before the worst outcomes occur, especially for vulnerable populations like the elderly, those with co-infections, or patients facing social and economic hardship.
Tamil Nadu Health Minister Ma. Subramanian, in a statement on Monday, praised the rollout and reaffirmed the government’s commitment to leveraging technology for public health impact. “We are proud to be the first state in India to use artificial intelligence in such a critical health domain. This is about giving every TB patient a better chance of survival, no matter where they live or how poor they are,” he said.
As the model is scaled across all districts in the state, Tamil Nadu hopes to set a precedent for evidence-based, tech-driven health solutions that are both scalable and equitable. Experts believe that if successful, this AI tool could play a key role in India’s broader TB elimination strategy and possibly be adapted for use in predicting other high-risk infectious diseases in the future.
Note to readers: This article is intended for informational purposes only and is not a substitute for professional medical advice, diagnosis, or treatment. Always consult with your doctor or a qualified healthcare provider before making changes to your health management plan, especially if you or someone you know is undergoing treatment for tuberculosis.
