COTS

COTS calculates consensus-based clinical scores to support decisions on initiating antitubercular therapy (ATT) for patients with ocular tuberculosis.

Key Features:

• Consensus derivation: Consensus recommendations were derived using a systematic two-step Delphi method involving 81 experts.
• Scenario evaluation: The model is based on assessment of 486 clinical scenario–based questions representing unique combinations of diagnostic components.
• Diagnostic components: Scenarios combine five elements—clinical phenotype, endemicity, tuberculin skin test, interferon-γ release assay, and chest X-Ray—with three clinical phenotypes further stratified by an additional component.
• Quantitative outputs: For each scenario the system reports a median score on a 1–5 scale and associated interquartile ranges (IQR).
• Encoded consensus table: Expert rankings were compiled into a consensus table and encoded to map input features to median scores.
• Score distribution: Strong consensus for initiating ATT (score 5) was observed in 114 of 486 scenarios (24%).

Scientific Applications:

• Treatment decision support: Provides consensus-derived scores to inform initiation of antitubercular therapy (ATT) in ocular tuberculosis cases.
• Clinical standardization: Translates COTS CON expert consensus into structured scores to standardize treatment-threshold assessments across phenotypes and endemicity.
• Research and validation: Supplies a quantitative reference of median scores and IQRs across 486 diagnostic scenarios for validation and outcome studies.

Methodology:

81 experts used a systematic two-step Delphi method to rank 486 clinical scenarios—each defined by clinical phenotype, endemicity, tuberculin skin test, interferon-γ release assay, and chest X-Ray (with three phenotypes further stratified); rankings were compiled into a consensus table reporting median scores and IQRs, and the table was encoded to map input feature combinations to a median score from 1 to 5.