LATENT TUBERCULOSIS SCREENING IN HIV-INFECTED WOMEN AT ANTENATAL CLINIC AND ANTI-RETROVIRAL THERAPY CENTER IN AN URBAN HOSPITAL IN INDIA
Embargo until
2015-05-01
Date
2014-04-24
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Johns Hopkins University
Abstract
Background: HIV-infected women have increased risk of reactivation of latent TB infection, making screening crucial. There are limited guidelines for latent TB screening in India. There is need to study the two available tests, which are known to be compromised in this population - Tuberculin Skin test (TST) and Quantiferon Gold-in-Tube assay (QGIT). Thus, our aim was to evaluate the performance of these two tests in two facilities – Antenatal Clinic (ANC) and Antiretroviral Therapy Center (ARTC).
Methodology: We conducted a cross-sectional study of HIV-infected women seeking care at the ANC and ARTC within a government hospital in Pune, India. Trained staff administered sociodemographic/medical history questionnaires before TST and QGIT testing. The percent positivity, percent agreement between tests results and kappa statistic, were calculated. Multivariate logistic regression was used to study risk factors for TST and QGIT positivity. Finally for each test, the number needed to screen (NNS) inorder to avert one TB case, was estimated in each facility.
Results: Among ANC participants, 15 (11%) were TST+ and 40 (29 %) were QGIT+, with 73.2 % agreement (95% CI = 63.2-81.4, k = 0.214 (0.04 - 0.39)). Among ARTC participants, 22 (15.1 %) were TST+ and 69 (47.2 %) were QGIT+, with 62.1 % agreement (95% CI = 52-71, k=0.22 (0.09-0.34)). TST-/QGIT+ discordance was common in both groups (21% ANC, 34% ARTC). It was observed in the ANC that NNS estimate is 200 for TST alone and 73 for QGIT alone. Similarly in ARTC TST alone has an NNS of 147 compared to 47 for QGIT alone.
Conclusions: The QGIT estimated a higher prevalence of LTBI as compared to TST in both facilities, more comparable to the Revised National TB Control Programme estimate of 30-40 %. Further advantage of using QGIT is the less than half of NNS estimate of TST alone in both settings. Thus there is need for development and evaluation of an algorithm with QGIT alone or two test strategy with both TST and QGIT in these settings, considering the issues with both the tests in this population.
Description
Keywords
Latent Tuberculosis infection, HIV infection, Pregnancy, Antenatal Clinic, Anti-retroviral Therapy Center, Tuberculin Skin Test, Interferon gamma release assay