A cross-sectional household contact study of smear-positive tuberculosis index cases was conducted in Karonga, Malawi, between Jan 1, 2013, and Dec 31, 2014. Prevalence of tuberculin skin test (TST) positivity was compared between household contacts aged 11 years of index cases (HIV-negative, HIV-positive not on ART, and HIV-positive on ART [HIV status at diagnosis of tuberculosis]). A positive TST was defined as ≥10 mm induration. Data were analysed with a random-effects logistic regression model to account for clustering.
There were 376 child contacts (148 index cases), of whom 330 (88%) had a TST reading. The proportion of contacts of index cases with the highest mycobacillary burden, with smear-positivity as a proxy, was similar: HIV-negative (78%, 149/191) and HIV-positive index cases on ART (71%, 50/70), compared with HIV-positive index cases not on ART (57%, 39/69). The odds of a positive TST were 60% lower in contacts of HIV-positive patients not on ART (adjusted odds ratio 0·4, 95% CI 0·1–1·0; p=0·04) and 70% lower in contacts of HIV-positive patients on ART (0·3, 0·1–0·8; p=0·02) than in HIV-negative index cases, adjusted for degree of contact and age of index case. The reduction in odds of TST positivity among contacts of HIV-positive patients on ART remained after additionally adjusting for degree of smear-positivity.
Contrary to expectation, we found a reduced prevalence of M tuberculosis infection among contacts of HIV-positive patients with tuberculosis on ART, after adjusting for degree of smear-positivity. This finding is suggestive of decreased transmission from these index cases. This reduction in transmission might be due to earlier diagnosis of tuberculosis in patients who are already engaged in care.
Wellcome Trust.