Predictors for identifying the most infectious pulmonary tuberculosis patient

Area of Science:

• Pulmonary Medicine
• Infectious Diseases
• Radiology

Background:

• Early diagnosis and management of tuberculosis (TB) are crucial for infection control.
• Clinicians require tools to assess the risk of infectious tuberculosis (TB) in patients with suspected disease.
• Acid-fast bacilli (AFB) smear results are critical for initiating isolation and empiric therapy.

Purpose of the Study:

• To identify demographic, clinical, and radiographic characteristics associated with AFB smear-positive TB.
• To develop a prediction rule for clinicians to estimate the risk of smear-positive TB.

Main Methods:

• Retrospective study comparing 105 AFB smear-positive TB patients with 52 AFB smear-negative TB patients.
• Data collected included demographic, clinical, and radiographic information.
• Multivariate analysis was used to identify independent predictors of smear-positive TB.

Main Results:

• Younger age, alcoholism, weight loss, fever, and specific radiographic findings (consolidation, infiltration, upper lung field lesions, cavitary pattern) were associated with smear-positive TB.
• Age < 65, fever, and upper lung field/consolidation/cavitary lesions were independent predictors.
• The prediction model achieved an area of 0.788 under the ROC curve.

Conclusions:

• A prediction model incorporating clinical and radiographic factors can aid in identifying patients likely to have smear-positive TB.
• This model can assist clinicians in making timely decisions regarding isolation and empiric anti-tuberculous therapy.
• The developed prediction rule offers a valuable tool for risk stratification in suspected TB cases.