Hidden TB: Most household cases show no symptoms

Tuberculosis (TB) can be hard to find because many people with the disease have no symptoms. More than half of TB found in community studies is classified as asymptomatic, meaning people do not report cough, fever, weight loss, or night sweats. To see how well common screening methods work in people who live with someone diagnosed with pulmonary TB, researchers led by Simon C. Mendelsohn enrolled adult household contacts in South Africa. The team looked for TB using three routes: asking about any TB symptoms, taking a chest radiograph (CXR) and checking it for any abnormality, and testing sputum in the lab. They applied sputum microscopy, Xpert Ultra, and liquid culture to confirm infection, and also measured C-reactive protein (CRP) in blood using a multiplex bead array to see if inflammation testing could help. By studying this specific group—people who had lived with a known TB patient—the researchers aimed to evaluate how many infections would be missed if screening relied only on symptoms or CXR, approaches commonly used in public health surveys.

Between March 2021 and December 2022, the study enrolled 979 household contacts aged 18 or older at three sites. Of these volunteers, 185 (18.9%) were living with HIV and 187 (19.1%) reported previous TB. The team performed systematic symptom screening (any duration), CXR (any abnormality), sputum microscopy, Xpert Ultra, and liquid culture. Serum C-reactive protein (CRP) was measured by multiplex bead array. Prevalent TB was defined as microbiologically-confirmed disease by Xpert Ultra or culture. Overall, 51 participants (5.2%) had prevalent TB, and 42 of those 51 (82.4%) were asymptomatic. Only 13 of the 42 asymptomatic cases (31.0%) were smear-positive; among those 13 smear-positive samples, 8/13 (61.5%) were graded scanty or 1+. CRP did not distinguish healthy contacts from those with asymptomatic TB (AUC 0.60; 95%CI 0.47-0.73). An abnormal CXR was seen in 23 of 41 asymptomatic cases (sensitivity 56.1%, 95%CI 41.0-70.1%), compared with 8 of 9 symptomatic cases (sensitivity 88.9%, 95%CI 56.5-98.0%). Combining symptom screening with CXR produced a sensitivity of 64.0% (32/50, 95%CI 50.1-75.9%) for detecting all prevalent TB.

The study’s findings are striking for programs that rely on symptoms and chest X-rays to find TB. More than 80% of confirmed TB among household contacts was asymptomatic, and CXR failed to flag over 40% of these asymptomatic infections. CRP testing did not effectively separate people with asymptomatic TB from healthy contacts. These results suggest that symptom- and CXR-based approaches fall short of the WHO Target Product Profile for a TB screening test, which specifies a minimum of 90% sensitivity and 70% specificity. For national TB prevalence surveys and routine screening in high-burden settings, relying on symptoms and CXR without universal sputum microbiological testing risks a substantial undercount of infections. The authors conclude that surveys and screening programs should consider broader use of microbiological testing such as Xpert Ultra and culture to better detect hidden TB and to more accurately estimate the burden of disease in endemic countries.