Tamil Nadu has become the country's first state to adopt a model that predicts the likelihood of adult Tuberculosis patients dying and has integrated it into its present state-wide use TB SeWA, which triages them upon diagnosis. The predictive model, developed by ICMR's National Institute of Epidemiology (NIE) and released last week, will reduce the average duration of diagnosis to hospitalisation for critically ill TB patients, thereby reducing fatality rates even further, said Dr Asha Frederick, Tamil Nadu State TB Officer.
The feature has been incorporated in Tamil Nadu's current TB SeWA (Severe TB Web Application), which was first implemented in 2022 under the state initiative of differentiated care model Tamil Nadu - Kasanoi Erappila Thittam (TN-KET), she informed the media. In TN-KET, health workers screen a triage every new TB-diagnosed adult for very severe undernutrition, respiratory distress or poor physical condition on five variables — body mass index (BMI), pedal oedema (swelling of feet and ankles), respiratory rate, oxygen saturation, and being able to stand without support.
Secondly, the health staff enters these variables into TB SeWA, and it will indicate if a specific patient is badly ill or not. In TN-KET, all the severely ill (very severely undernourished or with respiratory distress or poor physical status) must be given priority for admission, Dr Frederick explained.
So far, TB SeWA identified patients as 'severely ill' according to these parameters, informing health staff to prioritise them for inpatient services, Director of NIE Dr Manoj Murhekar stated. "The new feature does one more step" calculating and displaying projected probability of death of adults with TB, Dr Murhekar stated.
Also read: TN language rights war raging like whirlwind in Maha: CM Stalin
This goal risk percentage was designed to break down any subjective inference for severity and prompt frontline staff to take immediate and strict action for the hospitalisation of severely ill adults with TB upon diagnosis, he added.
"How this feature extension is useful is that the estimated probability of death ranges widely from a 'severely ill' to 'not severely ill' patient. The estimated probability for a severely ill adult with TB death ranges from 10 per cent to as much as 50 per cent, depending on the number of conditions that are present.".
"In sharp contrast, for patients who are not 'severely ill', the likelihood drops to a mere 1-4 per cent," said Dr Hemant Shewade, a senior Scientist at NIE. Statistics of the last three years indicate that nearly 10-15 per cent of adults with TB in Tamil Nadu are severely ill at diagnosis, he informed the media. "This objective risk estimate will ensure that the sickest patients are treated in hospitals without any delay," Dr Shewade said.
Adding further, Dr Shewade informed that though the average time from diagnosis to hospital admission of a very sick patient under TN-KET is one day, about a quarter of very sick patients continue to have delays of three to six days in the state. "There are still a few very sick patients — such as 25 per cent approximately who are admitted after a delay. This feature will appeal and encourage the healthcare personnel to take instant decisions on their referral for admission," he added.
"Over the period it will also enable us to analyse if the average time from diagnosis to admission for a very sick TB patient has decreased or not and soon this will help reduce TB deaths further in the state. Two-thirds of all TB deaths within two months of diagnosis," he added.
The model for predicting deaths has been created by taking data from nearly 56,000 TB patients who were treated in public health institutions in Tamil Nadu from July 2022 to June 2023.
Dr Frederick added that in predicting TB deaths, the five variables of triage that were being used under TN-KET alone were as good as all the baseline variables that were tracked through India's national TB portal Ni-kshay. Dr Shewade additionally added that all the baseline variables in Ni-kshay are accessible only by approximately three weeks (too delayed to utilise for prediction) whereas five-triage variables are captured within a day in Tamil Nadu.
All 2,800 public health institutions of Tamil Nadu — from primary health centres to medical colleges — are now using the TB SeWA application in addition to a paper-based triage tool, Dr Frederick said. "Tamil Nadu is thus the sole state in India to systematically document and utilise these five triage variables to inform patient management," she stated.
In accordance with an NIE study, since executing TN-KET for about three years now, losses in the care cascade have come down considerably and about two-third districts have reported decline in TB death rates. The ICMR-NIE researchers pointed out that the programme is a good example for other states too, where deaths from TB, particularly the early ones, continue to be an ongoing problem despite providing free diagnosis and treatment.
Follow us
