Predictors of latent tuberculosis treatment initiation and completion at a U.S. public health clinic: a prospective cohort study.

BACKGROUND: Treatment of latent tuberculosis infection (LTBI) is a key component in U.S. tuberculosis control, assisted by recent improvements in LTBI diagnostics and therapeutic regimens. Effectiveness of LTBI therapy, however, is limited by patients' willingness to both initiate and complete treatment. We aimed to evaluate the demographic, medical, behavioral, attitude-based, and geographic factors associated with LTBI treatment initiation and completion of persons presenting with LTBI to a public health tuberculosis clinic. METHODS: Data for this prospective cohort study were collected from structured patient interviews, self-administered questionnaires, clinic intake forms, and U.S. census data. All adults (>17 years) who met CDC guidelines for LTBI treatment between January 11, 2008 and May 6, 2009 at Wake County Health and Human Services Tuberculosis Clinic in Raleigh, North Carolina were included in the study. In addition to traditional social and behavioral factors, a three-level medical risk variable (low, moderate, high), based on risk factors for both progression to and transmission of active tuberculosis, was included for analysis. Clinic distance and neighborhood poverty level, based on percent residents living below poverty level in a person's zip code, were also analyzed. Variables with a significance level <0.10 by univariate analysis were included in log binomial models with backward elimination. Models were used to estimate risk ratios for two primary outcomes: (1) LTBI therapy initiation (picking up one month's medication) and (2) therapy completion (picking up nine months INH therapy or four months rifampin monthly). RESULTS: 496 persons completed medical interviews and questionnaires addressing social factors and attitudes toward LTBI treatment. 26% persons initiated LTBI therapy and 53% of those initiating completed therapy. Treatment initiation predictors included: a non-employment reason for screening (RR 1.6, 95% CI 1.0-2.5), close contact to an infectious TB case (RR 2.5, 95% CI 1.8-3.6), regular primary care(RR 1.4, 95% CI 1.0-2.0), and history of incarceration (RR 1.7, 95% CI 1.0-2.8). Persons in the "high" risk category for progression/transmission of TB disease had higher likelihood of treatment initiation (p < 0.01), but not completion, than those with lower risk. CONCLUSIONS: Investment in social support and access to regular primary care may lead to increased LTBI therapy adherence in high-risk populations.