Tamil Nadu Launches Predictive Model for TB Mortality
Tamil Nadu has become the first state of India to use a predictive model for estimating the risk of death among adults diagnosed with Tuberculosis (TB). This model is integrated with the state’s existing TB SeWA application, which assists healthcare workers in triaging patients immediately at diagnosis. The initiative aims to reduce delays in hospital admission for severely ill TB patients and lower mortality rates.
Background and Recent Developments
The predictive model was developed by the Indian Council of Medical Research’s National Institute of Epidemiology (ICMR-NIE). It was launched recently and incorporated into Tamil Nadu’s TB SeWA (Severe TB Web Application). TB SeWA has been operational since 2022 under the Tamil Nadu Kasanoi Erappila Thittam (TN-KET), a differentiated care model focusing on early identification of severe TB cases.
Screening and Triage Process
Under TN-KET, healthcare workers screen newly diagnosed adult TB patients for severe illness using five key variables – body mass index (BMI), pedal oedema, respiratory rate, oxygen saturation, and ability to stand unaided. These variables are entered into TB SeWA, which flags patients as severely ill or not. Those flagged severely ill are prioritised for immediate hospital admission.
Predictive Model and Risk Assessment
The new feature in TB SeWA calculates the predicted probability of death for each patient. This risk percentage ranges from 10% to 50% for severely ill patients depending on the number of critical conditions present. For patients not flagged as severely ill, the risk drops sharply to 1-4%. This objective risk score helps frontline staff make prompt decisions, reducing subjective errors in assessing severity.
Impact on Patient Care and Mortality
Data from the past three years show that 10-15% of adult TB patients in Tamil Nadu are severely ill at diagnosis. The average time from diagnosis to admission for these patients is one day, but about 25% still face delays of up to six days. The predictive model aims to eliminate these delays and ensure timely hospitalisation, which is critical since two-thirds of TB deaths occur within two months of diagnosis.
Data and Coverage
The model was created using data from nearly 56,000 TB patients diagnosed in public health facilities across Tamil Nadu between July 2022 and June 2023. The five triage variables used are as accurate in predicting TB deaths as the full set of baseline variables in India’s national TB portal, Ni-kshay. However, Ni-kshay data becomes available only after three weeks, whereas Tamil Nadu’s triage variables are recorded within a day, enabling faster action.
Statewide Implementation
All 2,800 public health facilities in Tamil Nadu, from primary health centres to medical colleges, use the TB SeWA application alongside a paper-based triage tool. Tamil Nadu remains the only state in India to systematically collect and use these five triage variables to guide TB patient management.
Significance and Future Prospects
The TN-KET initiative has reduced losses in the TB care cascade and helped many districts lower TB death rates. It serves as a benchmark for other states struggling with early TB fatalities despite free diagnosis and treatment. The predictive model’s ongoing use will be monitored to assess reductions in admission delays and mortality.
| Related | |
|---|---|
