Why are States asked to use untested AI tool for TB screening?

Why are States asked to use untested AI tool for TB screening?

Sub : Sci

Sec: Health

Context

• The Central TB Division (CTD) has informed States to “consider utilising” the DeepCXR tool for AI-assisted chest X-ray interpretation.
• The recommendation comes despite the availability of two indigenously developed AI solutions (qXR and Genki) that have undergone Health Technology Assessment (HTA).
• The move raises concerns about the transparency and effectiveness of AI adoption in TB screening.

HTA Assessment and Approval Process

• HTA assessment is not mandatory but is usually awaited before implementing new technology in the TB programme.
• Example: TrueNat (TB diagnosis) and BPaLM/BPaL regimen (MDR-TB treatment) were assessed by HTA before inclusion in the TB programme.
• The qXR tool (Qure.ai) and Genki (DeepTek) were assessed and approved by HTA and the Medical Technology Assessment Board (MTAB), yet not included in programmatic implementation.

Issue with DeepCXR Implementation

• DeepCXR, developed by the Institute for Plasma Research, Gandhinagar, was approved by an ICMR expert committee without HTA assessment.
• Unlike qXR and Genki, DeepCXR lacks published data on sensitivity, specificity, and field performance.
• No official communication from CTD to States about its recommendation; States were informed only when they requested AI solutions.

Comparison of AI Tools for TB Screening

Importance of AI-Assisted Chest X-Ray Interpretation

• Crucial for screening presumptive and subclinical TB cases.
• National TB Prevalence Survey (2019-2021): 42.6% cases detected via chest X-ray.
• Tamil Nadu TB Prevalence Survey (2021-2022): 39% cases detected through chest X-ray.
• AI interpretation provides quick, accurate, and cost-effective TB screening, beneficial for resource-limited settings.

National Tuberculosis Elimination Programme (NTEP)

Introduction:

• Renaming: In 2020, the Revised National Tuberculosis Control Programme (RNTCP) was renamed as the National TB Elimination Program (NTEP) to emphasize India’s commitment to eliminate TB by 2025, five years ahead of the global target of 2030.

Sustainable Development Goals (SDG) Targets (Baseline 2015):

• 80% reduction in TB incidence.
• 90% reduction in TB mortality.
• Zero TB patients and their households facing catastrophic costs due to TB.

Strategic Pillars: The program operates under four strategic pillars: Detect, Treat, Prevent, and Build (DTPB).

Estimated TB Burden in India (as per Global TB Report 2023)

• Incidence: 2.82 million new TB cases in 2022, equating to 199 cases per 100,000 population.
• Mortality: 331,000 deaths in 2022, or 23 deaths per 100,000 population.
• HIV Co-infection: Approximately 2% of TB patients are estimated to be HIV positive.
• Drug-Resistant TB (DR-TB): 2.5% in new cases and 13% in previously treated cases.
• Global Context: India accounts for nearly 27% of the global TB incidence, with 2.8 million of the estimated 10.6 million global TB cases in 2022 occurring in India.