RESUMO
BACKGROUND: Close contacts of patients with tuberculosis (TB) have a substantial risk of developing the disease, particularly during the first year after exposure. Household contact investigation has recently been recommended as a strategy to enhance case detection in high-burden countries. However the barriers to its implementation in these settings remain poorly understood. METHODS: A nested case-control study was conducted in Vietnam within the context of a large cluster randomised controlled trial of active screening for TB in household contacts of patients with pulmonary TB. The study population comprised contacts (and their index patients) from 12 Districts in six provinces throughout the country. Cases were contacts (and their index patients) that did not attend the scheduled screening appointment. Controls were those who did attend. We assessed relevant knowledge, attitudes and practices in cases and controls. RESULTS: The acceptability of contact investigation was high among both cases (n = 109) and controls (n = 194). Both cases (47%) and controls (36%) commonly reported discrimination against people with TB. Cases were less likely than controls to understand that sharing sleeping quarters with a TB patient increased their risk of disease (OR 0.46, 0.27 - 0.78) or recognise TB as an infectious disease (OR 0.65, 0.39 - 1.08). A higher proportion of cases than controls held the mistaken traditional belief that a non-infectious form of TB caused the disease (OR 1.69, 1.02 - 2.78). CONCLUSIONS: The knowledge, attitudes and practices of contacts and TB patients influence their ongoing participation in contact investigation. TB case detection policies in high-prevalence settings can be strengthened by systematically evaluating and addressing locally important barriers to attendance. TRIAL REGISTRATION: Australian New Zealand Clinical Trials Registry, ACTRN12610000600044 .