Open Access Open Badges Research article

An evaluation of factors associated with taking and responding positive to the tuberculin skin test in individuals with HIV/AIDS

Líbia CRV Moura1*, Ricardo AA Ximenes12, Heloísa L Ramos1, Demócrito B Miranda Filho2, Carolina DP Freitas1, Rosangela MS Silva3, Isabella Coimbra1, Joanna d'Arc L Batista3, Ulisses R Montarroyos1 and Maria de Fátima P Militão Albuquerque3

Author Affiliations

1 Department of Tropical Medicine, Universidade Federal de Pernambuco, Recife, Brazil

2 Department of Medical Science, Universidade de Pernambuco, Recife, Brazil

3 NESC Department, Centro de Pesquisas Aggeu Magalhães/FIOCRUZ, Recife, Brazil

For all author emails, please log on.

BMC Public Health 2011, 11:687  doi:10.1186/1471-2458-11-687

Published: 5 September 2011



The tuberculin skin test (TST) is still the standard test for detecting latent infection by M tuberculosis (LTBI). Given that the Brazilian Health Ministry recommends that the treatment of latent tuberculosis (LTBI) should be guided by the TST results, the present study sets out to describe the coverage of administering the TST in people living with HIV at two referral health centers in the city of Recife, where TST is offered to all patients. In addition, factors associated with the non-application of the test and with positive TST results were also analyzed.


A cross-sectional study was carried out with HIV patients, aged 18 years or over, attending outpatient clinics at the Correia Picanço Hospital/SES/PE and the Oswaldo Cruz/UPE University Hospital, who had been recommended to take the TST, in the period between November 2007 and February 2010. Univariate and multivariate logistic regression analyses were carried out to establish associations between the dependent variable - taking the TST (yes/no), at a first stage analysis, and the independent variables, followed by a second stage analysis considering a positive TST as the dependent variable. The odds ratio was calculated as the measure of association and the confidence interval (CI) at 95% as the measure of accuracy of the estimate.


Of the 2,290 patients recruited, 1087 (47.5%) took the TST. Of the 1,087 patients who took the tuberculin skin test, the prevalence of TST ≥ 5 mm was 21.6% among patients with CD4 ≥ 200 and 9.49% among those with CD4 < 200 (p = 0.002). The patients most likely not to take the test were: men, people aged under 39 years, people with low educational levels and crack users. The risk for not taking the TST was statiscally different for health service. Patients who presented better immunity (CD4 ≥ 200) were more than two and a half times more likely to test positive that those with higher levels of immunodeficiency (CD4 < 200).


Considering that the TST is recommended by the Brazilian health authorities, coverage for taking the test was very low. The most serious implication of this is that LTBI treatment was not carried out for the unidentified TST-positive patients, who may consequently go on to develop TB and eventually die.