Spotting the old foe—revisiting the case definition for TB

Abstract

Disease case definitions are important instruments for clinical care, interventional research, and surveillance. Therefore, it is concerning that the current case definitions for tuberculosis remain underscored by the classic paradigm of binary states of latent infection and active disease, with a stepwise, linear transition under which symptoms, bacteriological positivity, and disease pathology are assumed to emerge broadly together (figure, A).1 This assumption has resulted in a reliance on symptom screening to distinguish these two states. However, in recent prevalence surveys, 40–79% of bacteriologically positive tuberculosis occurs in the absence of patient-recognised tuberculosis symptoms.2 Rather than explicitly addressing this discordance, tuberculosis case definitions are often ambiguous regarding tuberculosis symptoms, or internally inconsistent.