Exhaled breath analysis: A promising triage test for tuberculosis in young children

Area of Science:

• Pediatric pulmonology
• Infectious disease diagnostics
• Biomarker discovery

Background:

• Diagnosing pediatric pulmonary tuberculosis, particularly in non-expectorating infants, presents significant clinical challenges.
• Exhaled breath analysis offers a non-invasive method for respiratory infection diagnosis, but its application in pediatric tuberculosis requires further investigation.
• Limited data exists on the efficacy of breath testing for diagnosing tuberculosis in children under five years old.

Purpose of the Study:

• To evaluate the diagnostic accuracy of exhaled breath analysis using a novel device in children under five years with suspected tuberculosis.
• To assess the potential of breath analysis as a triage tool for pediatric tuberculosis in a resource-limited setting.

Main Methods:

• A prospective cross-sectional study was conducted in Kenya involving 118 children under five years presenting with tuberculosis symptoms.
• Exhaled breath samples were collected using a portable, battery-powered device.
• Machine learning algorithms were employed to analyze breath samples, classifying them as microbiologically confirmed tuberculosis (positive) or unlikely tuberculosis (negative).

Main Results:

• The optimal machine learning model achieved an area under the curve (AUC) of 0.73.
• The test demonstrated a sensitivity of 86% (95% CI 62-96%) and a specificity of 42% (95% CI 30-55%).
• While promising, the performance metrics did not meet the World Health Organization's target product profile for tuberculosis diagnostics.

Conclusions:

• Exhaled breath analysis, coupled with machine learning, shows potential as a non-invasive screening or triage test for pediatric tuberculosis.
• Further research and technological refinement are necessary to improve specificity and meet established diagnostic targets for widespread clinical adoption.
• This approach could facilitate earlier detection and management of tuberculosis in young children, especially in settings with limited diagnostic infrastructure.