A NOVEL clinical decision support system (CDSS), TBorNotTB, has been developed to assist clinicians in diagnosing pulmonary tuberculosis (TB) in low-prevalence settings, potentially reducing diagnostic delays and improving patient care.
The system was designed using a modified Delphi method and incorporates factors such as epidemiologic risk, TB history, symptoms, chest imaging, and sputum/bronchoscopy results to guide diagnosis and management decisions. The CDSS works by assigning points based on these factors, and when the total score falls below a predefined threshold, it automatically discontinues airborne isolation precautions. However, if the score suggests a higher likelihood of TB, additional evaluation, including a review of infection control measures, is recommended.
To validate the system, researchers applied the TBorNotTB tool retrospectively to hospitalised patients within the Mass General Brigham system from July 2016–December 2022. A total of 104 individuals with culture-confirmed pulmonary TB cases and an equal number of age-matched controls with negative mycobacterial cultures were included in the analysis. Key predictors of TB identified included prior residence in a highly endemic country, weight loss, positive interferon release assay, lack of symptom resolution with treatment for alternative diagnoses, and TB-suggestive chest imaging.
The final version of the CDSS demonstrated an impressive 100% sensitivity for detecting TB, even in cases with negative acid-fast bacillus smears. However, it had a more modest specificity of 27%, with an area under the curve of 0.87. These results suggest that the tool is highly effective at identifying individuals who need further evaluation, although there is still a risk of false positives.
This CDSS, which is integrated into the electronic medical record system, could significantly reduce unnecessary airborne isolation, minimise time spent reviewing suspected TB cases, and lower the risk of nosocomial transmission. The tool’s development represents an important step toward improving TB diagnosis and management in healthcare settings with low TB prevalence.
Ada Enesco, EMJ
Reference
Dugdale CM et al. TB or not TB? Development and validation of a clinical decision support system to inform airborne isolation requirements in the evaluation of suspected tuberculosis. Infect Control Hosp Epidemiol. 2025; DOI:10.1017/ice.2024.214.