Predictors of failure in timely tuberculosis treatment completion, United States

Abstract

Setting: The US tuberculosis (TB) surveillance system.

Objective: To examine failure in timely TB treatment completion to identify interventions toward achieving the national goal of ≥ 93% treatment completion in ≤ 12 months among patients eligible for 6-9 month regimens.

Design: We examined 1993-2006 trends in timely treatment completion; for 2006 cases, we used Poisson regression to assess predictors for failure in timely completion.

Results: Timely treatment completion improved from 64% in 1993 to 84% in 2006, with similar trends among foreign- and US-born persons and racial/ethnic subgroups. Annual increases in timely completion were ≤ 1 percentage point during 1998-2006. Subpopulations at highest risk for failure in timely completion were persons with combined pulmonary and extra-pulmonary disease (foreign-born adjusted RR [aRR] 3.25, 95%CI 2.47-4.28; US-born aRR 2.75, 95%CI 1.98-3.83) or incarceration (foreign-born aRR 2.30, 95%CI 1.80-2.93; US-born aRR 1.71, 95%CI 1.36-2.14). Homelessness and human immunodeficiency virus infection were other risk factors.

Conclusions: Particular attention to timely completion is needed for subpopulations requiring strong medical expertise in TB management and those at risk for treatment non-adherence, especially if foreign-born. Understanding and addressing causes of delayed completion and improving documentation of treatment completion among all cases will be crucial to achieving the US goal.