Introduction: Tuberculosis (TB) is one of the most common infectious diseases worldwide. One third of the world’s population is infected with TB. Mycobacterium tuberculosis, the bacterium that causes TB, can infect the body and remain inactive (latent infection), or the infection can progress and cause the disease. Certain populations are at a particular risk to develop TB; these groups include persons infected with HIV, homeless persons, incarcerated individuals, and those who live in areas where TB is endemic, such as Southeast Asia. The standard treatment regimen to prevent progression to TB is the nine months of self-administered isoniazid (300 mg) (also known as the 9H regimen) for the treatment of latent TB infection (LTBI). However, a short regimen of 12 weekly doses of isoniazid (900 mg) plus rifapentine (900 mg), (also known as the 3HP regimen) has been recently shown to be as effective as the standard regimen with higher treatment completion rates by the PREVENT TB trial.

The objective of this analysis is to evaluate factors associated with non-completion of LTBI treatment among homeless participants enrolled in the PREVENT TB trial.

Methods: The study population consisted of 6, 232 participants enrolled in the PREVENT TB trial after excluding participants enrolled in non-North American sites, pregnant women, and children. Non-completion of treatment was defined as receipt of: ≥1 of the first 3 observed doses in the 3HP regimen followed by receiving an observed dose during the treatment period; or missing ≥1 of the first 3 monthly visits in the 9H regimen, followed by a monthly visit at any time during treatment. Chi-square test, univariate and multivariate logistic regression analyses were conducted to evaluate demographic, clinical, social, and behavioral factors associated with non-completion of LTBI treatment in the homeless population by using SAS^® version 9.3 (Cary, North Carolina).

Results: The cohort for analysis included 6,232 participants (3HP = 3,230 and 9H=3,002), of which 505 were homeless (8.1%). Most homeless participants were male (86.7%), born in the U.S. (87.1%), and African descendants (53.1%). The median age was 46 compared with 36 in the non-homeless group. Other characteristics associated with homelessness were unemployment, smoking, alcohol consumption, and use of concomitant medications. Homeless participants in both treatment regimens were more likely to be of African descent (p

Conclusion: Homeless persons present many challenges in treatment for any disease, particularly one that requires lengthy treatment and follow-up such as tuberculosis, as well as prophylactic treatment for LTBI. Appropriate interventions to improve completion of LTBI treatment such as coordination with social and educational programs, as well as with homeless shelters, could increase adherence among the homeless population.