Integrated treatment-decision algorithms for childhood TB: modelling diagnostic performance and costs

Area of Science:

• Public Health
• Infectious Disease Diagnostics
• Health Economics

Background:

• Childhood tuberculosis (TB) diagnosis requires improved tools for children under 10.
• Treatment-decision algorithms (TDAs) were developed with and without chest X-ray (CXR) to aid diagnosis.
• Implementation cost and diagnostic performance of TDAs in Ugandan primary health centers (PHCs) and district hospitals (DHs) were modeled.

Purpose of the Study:

• To model the diagnostic performance and costs of implementing TDAs in Ugandan PHC and DH settings.
• To compare various diagnostic testing combinations within TDAs.
• To evaluate the cost-effectiveness of different TDA implementation strategies for childhood TB.

Main Methods:

• Decision-tree models were used to simulate TDA pathways for diagnosing TB in 10,000 children.
• Six scenarios combining diagnostic tests (Xpert, urine LAM, CXR) were evaluated at PHCs and DHs.
• Monte Carlo simulation assessed diagnostic accuracy and cost per correct treatment-decision from a health system perspective.

Main Results:

• TDAs demonstrated high sensitivity (80.8%-91.9%) but low specificity (50.9%-60.9%) across all scenarios.
• Total costs ranged from I$1,768,958 to I$2,470,298, with overtreatment of false positives being a major cost driver.
• Mobile CXR at PHCs yielded the lowest cost per treatment-decision (I$287), while DH referral was highest (I$449).

Conclusions:

• TDAs offer high sensitivity for childhood TB diagnosis and are more cost-effective at PHCs than DHs.
• Improving specificity and reducing overtreatment costs are crucial for affordable, large-scale TDA implementation.
• Mobile CXR integrated into TDAs at PHCs presents a promising strategy for cost-effective childhood TB diagnosis.