Auld AF, Kerkhoff AD, Hanifa A, et al. - Researchers herein sought to develop a parsimonious multivariable prognostic model to better facilitate early tuberculosis (TB) diagnosis and safer scale-up of tuberculosis preventive therapy (TPT) to people living with HIV (PLHIV) by avoiding TPT prescription to clients with asymptomatic active TB, who need TB treatment.

• Botswana XPRES trial data for adult HIV clinic enrollees collected during 2012 to 2015 were used to derive a TB risk score for PLHIV; validation of the score was performed on 3 external datasets.

• High sensitivity and ability to correctly rule out TB (ie, high negative predictive value (NPV)) at key time points in care, such as HIV clinic enrollment and before TPT prescription, were prioritized.

• The derivation dataset yielded following factors for inclusion in the clinical score: male gender (1 point), ≥ 1 WHO TB symptom (7 points), smoking history (1 point), temperature >37.5°C (6 points), body mass index (BMI) < 18.5kg/m² (2 points), and severe anemia (hemoglobin < 8g/dL) (3 points).

• The 6 most important predictors, commonly available in LMIC clinic settings, were identified using both logistic regression and random forest machine learning approaches.

• In the external datasets, TB risk score ≥ 2 exhibited higher sensitivity (87% to 97%) when compared with WHO 4-symptom screening rule and increased NPV by 0.3% to 1.7%.

• The score led to identification of three risk groups, with active TB prevalence in external datasets ranging from 1% to 6% in the lowest to 22% to 32% in the highest risk groups.