Development and validation of a prediction model based on a nomogram for tuberculous pleural effusion

Area of Science:

• Pulmonology
• Infectious Diseases
• Medical Diagnostics

Background:

• Diagnosing tuberculous pleural effusion (TPE) presents significant clinical challenges.
• A notable gap exists in comparative analyses of TPE prediction models.
• Existing diagnostic tools for TPE require further validation and comparison.

Purpose of the Study:

• To develop and validate a novel prediction model for tuberculous pleural effusion (TPE).
• To compare the diagnostic performance of the new TPE model against existing prediction models.
• To enhance the accuracy and reliability of TPE diagnosis.

Main Methods:

• Development and validation of a TPE prediction model using training, testing, and external validation sets.
• Variable selection employed LASSO and logistic regression techniques.
• Performance evaluation included discriminability (AUC), calibration, and clinical utility (decision curve analysis).

Main Results:

• The novel TPE prediction model achieved high diagnostic accuracy with AUCs of 0.931, 0.856, and 0.925 across the validation sets.
• Key predictors included fever, interferon-gamma release assays, and pleural fluid biomarkers (ADA, LDH, CEA, CYFRA 21-1).
• The developed model demonstrated superior performance compared to two existing TPE prediction models (AUCs 0.793 and 0.854).

Conclusions:

• The newly developed TPE prediction model exhibits robust diagnostic performance.
• This model offers improved accuracy and clinical utility over existing TPE prediction tools.
• The findings support the use of this novel model for more effective TPE diagnosis.