In a significant step towards reducing tuberculosis (TB) fatalities, Tamil Nadu has become the first state in India to implement a predictive model that estimates the risk of death in adults diagnosed with the disease, right at the point of diagnosis.
Integrated into the state’s TB SeWA (Severe TB Web Application) system, the model was developed by the Indian Council of Medical Research’s National Institute of Epidemiology (ICMR-NIE) and launched last week. It is designed to help healthcare workers identify and prioritise critically ill patients for immediate hospitalisation.
Also Read | PM Modi calls for ‘whole of society, govt’ push to end TB in India by 2025
“This model calculates a patient’s probability of death based on clinical indicators like body mass index (BMI), pedal edema, respiratory rate, oxygen saturation, and the ability to stand unassisted,” said Dr. Asha Frederick, Tamil Nadu’s State TB Officer. These five variables are recorded under the state’s differentiated care initiative, Tamil Nadu – Kasanoi Erappila Thittam (TN-KET), operational since 2022.
Until now, the TB SeWA tool helped staff flag “severely ill” patients using the triage variables. With the newly added predictive feature, the tool now displays a death risk percentage to guide more immediate decisions. “This offers a clear, objective metric, replacing subjective judgement and accelerating referrals,” said Dr. Manoj Murhekar, Director of ICMR-NIE.
Analysis shows that the probability of death among patients flagged as severely ill can range from 10% to as high as 50%, compared to just 1–4% in those not flagged. “This sharp contrast motivates faster action,” explained Dr Hemant Shewade, senior scientist at NIE.
Currently, about 10–15% of TB patients in the state are identified as severely ill at diagnosis. While TN-KET has reduced the average time from diagnosis to hospitalisation to one day, nearly 25% of critical patients still face delays ranging from three to six days. The predictive model aims to address this lag.
The model, based on data from nearly 56,000 TB patients diagnosed between July 2022 and June 2023, is unique in using only the five triage variables, captured within a day, unlike the national Ni-kshay portal, which takes weeks to provide complete data.
Dr Frederick noted that all 2,800 public health facilities in Tamil Nadu are now using TB SeWA with this enhanced capability. “Tamil Nadu is the only state that systematically records and uses these five triage variables to guide TB patient management,” she said.
Early evaluations suggest that since TN-KET’s introduction, patient care losses have decreased, and two-thirds of districts have seen a decline in TB death rates.
