The tuberculin test is a diagnostic method for detecting latent tuberculosis (TB) infection, especially among disease contact cases. The objective of this study was to analyze the prevalence and evolution of Mycobacterium tuberculosis infection among TB contact cases.

A retrospective cohort study was performed in a reference center for TB. The study population consisted of 2,425 patients who underwent a tuberculin test from 2003 to 2010 and whose results indicated contact with individuals with TB. The data were collected from the registry book of the tuberculin tests, patient files and the Information System Records of Notification Grievance. To verify the evolution of TB, case records through September 2014 were consulted. Data were analyzed using the Statistical Package for the Social Sciences (SPSS). In all hypothesis tests, a significance level of 0.05 was used.

From the studied sample, 435 (17.9%) contacts did not return for reading. Among the 1,990 contacts that completed the test, the prevalence of latent TB infection was 35.4%. Of these positive cases, 50.6% were referred to treatment; the dropout rate was 42.5%. Among all of the contacts, the TB prevalence was 1.8%, from which 13.2% abandoned treatment.

The collected data indicate the need for more effective public policies to improve TB control, including administering tests that do not require a return visit for reading, enhancing contact tracing and encouraging actions that reinforce full treatment adherence.