Interim analysis showed efficacy less than 2.5%.

The purpose of this study is to determine if the isoniazid is effective in the prevention of tuberculosis in a prison population, exposed to the high endemicity of the disease.

Despite being a known disease of mankind over 9000 years, tuberculosis (TB) is still a major public health problem in developing countries, mainly due to so-called highly endemic sites, such as prisons.

It is Infectious disease, with airborne transmission, TB can present both the active or latent form. Despite the biological aspect of transmission, unhealthy environmental conditions (room without direct sunlight, poor ventilation and overcrowded) and individual factors (malnutrition, immunosuppression, use of alcohol and other drugs) have significant influence on transferability and infectivity.

With the discovery of drugs active against Mycobacterium tuberculosis, it was observed a reduction of disease incidence in the world. Despite this declining incidence, World Health Organization (WHO) classifies TB as a public health problem due to the emergence of multidrug strains or extensively resistant to treatment, added to the cases of latent TB reactivation, observed with the advent of HIV/AIDS.

Studies show a direct relationship between the incidence of TB and the prison environment. Nevertheless, the presence of prisons in one location increases the incidence of this disease, indicating that there is an exchange of disease between the prison and the community.

Currently, control of TB in the prison system is based on the tracking of individuals with active TB and / or latent and, in their respective treatment. For the identification of individuals with active disease, it is necessary the recognition of respiratory symptoms (cough mainly) and sputum smear microscopy and sputum culture, and chest X-ray. As the search of individuals with latent form, must be carried to the skin test with Purified Protein Derived (PPD).

Most of the units of the Brazilian's prison system these diagnostic methods are not available and hence the transport of individuals for their realization is necessary, generating an additional cost, in most cases, unfeasible process execution.

With the completion of this study, it seek to determine the effectiveness of primary prophylaxis in the prison population in order to gather new scientific evidence, to bring affordable methods for the control of TB in prisons.

Despite advances in diagnosis and treatment TB, this is the third leading cause of death from infectious diseases worldwide (Naghavi et al, 2015). In 2015, the WHO estimated incidence of 9.6 million new cases of TB in the world, with about 1.1 million deaths. For Brazil it was estimated incidence of 44 cases per 100,000 (WHO, 2015). The incidence of TB has declined about 2% per year, but this rate is not homogeneous in the global context.

Brazil occupies the 22nd place in the ranking of the WHO with an estimated annual rate of 83,310 cases of the disease (Zumla et al, 2015). In the past seven years, it is estimated that the incidence declined only 0.7% (per year). A key factor in this slow progress in TB control in Brazil, and other emerging countries is the existence of high-risk subpopulations, including slums and prisons, which act as reservoirs and amplifiers for the transmission of the disease (Basu; Stückler; Mckee 2011; Dowdy et al, 2012). A recent systematic review showed that the average incidence of TB in the prison population can be up to 23 times that recorded in the general population (Baussano et al, 2010).

With the fourth largest prison populations in the world, is observed in Brazil, the increased incidence of TB among prisoners in the last seven years. Although prisoners represent only 0.3% of the population, the increase in the prison population over this period resulted in almost doubling the proportion of all TB cases that occur among prisoners (4.1% in 2007, 8, 1% in 2013).

The arrests are in an ideal environment for the spread of TB, since they show individuals users of tobacco and alcohol in high doses, in addition to drug abuse in overcrowded cells with poor ventilation and with limited access to care health and diagnosis of TB. Currently, the Ministry of Health recommends active search for TB at the entrance to the prison and once a year by chest X-ray. Due to the cost and logistics, most prisons do not adhere to this recommendation. There is also a clear recommendation not to use the tuberculin skin test or perform the treatment for latent tuberculosis. If the procedures for active case detection and / or prophylactic treatment would impact the high transmission in prisons is a question that still remains open (Al-Darraji; Kamarulzamn; Altice, 2012).

The concentration of cases of TB in prisons can represent both an obstacle and an opportunity to control the disease, depending on the effectiveness of interventions in these environments.

Preliminary studies show high annual rate of TB infection (26%) among the prison population of 12 penitentiaries of Mato Grosso do Sul. Besides the large burden of disease in this population is significant dispersion of the TB prison to the community (Sacchi et al, 2015). Cross-sectional studies show high yield annual screening for TB, however, the effectiveness of this measure combined with other interventions remains unclear (Ferreira et al, 1996; Fournet et al, 2006; Lemos et al, 2014; Vieira et al, 2010; Sanches et al, 2013; Walnut; Abrahão; Galesi, 2012; Kuhleis et al, 2012; Estevan; Oliveira; Croda, 2013). Due to the combination of high force of infection in prisons and short prison term, primary prophylaxis may be an effective intervention. This new approach has never been evaluated in the context of prisons in low and middle income countries; to assess the impact of program strategies for TB screening and prophylaxis, longitudinal data will be essential. Given the infrastructure that was created for long-term prospective studies in Brazil, there is an excellent opportunity to close the critical knowledge gaps that have been barriers to effective implementation of TB control in high transmission prisons.

Inclusion criteria:

• Age above 18 and under 45 at the time of inclusion;
• Sign the informed consent form.

Exclusion criteria:

• Be indigenous;
• Active TB or previous use of isoniazid;
• Score Alcohol Use Disorders Identification Test ≥15.
• Reactive serology for HIV, hepatitis B and C;
• Reactive result for quantiferon, considering as positive the result of Tube 1 and / or Tube22 above 0.2 IU / mL;
• Liver enzymes (Aspartate aminotransferase and Alanate aminotransferase) three times the upper limit;
• History or treatment for epilepsy;

All participants who received at least one dose of isoniazid or placebo.

9 months

On the intervention days, all participants were asked about the presence of any unusual signs and / or symptoms. Data were recorded in a log containing start and end date, event description, degree and intervention performed.

Loss to follow-up Interruption of the study Arms presented differential group allocation Statistical power to detect the primary endpoint was 73%, lower than estimated in our sample size calculation