A Measurement Model to Quantify Systemic Bacilli Load in Severe HIV-Associated Tuberculosis

Area of Science:

• Infectious Diseases
• Immunology
• Biostatistics

Background:

• Pathogen load is critical for understanding infection, but current measurement methods are often inadequate.
• Accurate quantification of mycobacterial load in HIV-associated tuberculosis is challenging.

Purpose of the Study:

• To develop and validate a novel method for estimating mycobacterial load in HIV-associated tuberculosis using latent variable modeling.
• To assess the correlation of this novel pathogen load estimation with host inflammation and mortality.

Main Methods:

• Latent variable modeling was applied to data from 519 inpatients with HIV-associated tuberculosis.
• Urine and blood tests, rather than sputum, were utilized to capture mycobacterial load.
• Measurement models were used to summarize covariance between different diagnostic tests.

Main Results:

• Substantial covariance was observed between urine and blood tests for mycobacterial detection.
• The estimated pathogen loads derived from latent variable modeling showed a stronger correlation with host inflammation markers.
• Estimated pathogen loads were also more strongly correlated with patient mortality compared to traditional measures like blood culture time to positivity.

Conclusions:

• Latent variable modeling provides a more robust estimation of mycobacterial load in HIV-associated tuberculosis.
• Urine and blood tests can serve as valuable, non-sputum-based indicators for pathogen load assessment.
• This improved pathogen load estimation has significant implications for understanding disease pathobiology and predicting patient outcomes.