BACKGROUND: The treatment of latent tuberculosis infection (LTBI) is essential for tuberculosis elimination in the United States, but the major limitation is poor adherence to therapy. To aid the design of targeted adherence interventions, we investigated the factors associated with noncompletion of isoniazid (INH) therapy for LTBI. METHODS: A retrospective analysis of patients with who failed to complete vs those who completed 9 months of INH therapy at the RISE TB Clinic (Miriam Hospital; Providence, RI) in 2003 was performed. Factors associated with treatment noncompletion were examined using univariate and multiple logistic regression analysis. RESULTS: Of 845 patients with LTBI, 690 patients (81.6%) initiated INH therapy, of whom 426 patients (61.7%) completed therapy, and 246 patients (35.6%) were lost to follow-up. Treatment was discontinued in 18 patients (2.6%). Patients who failed to complete therapy were younger (mean age, 30.6 vs 33.8 years, respectively; p = 0.006), and were more likely to be uninsured (42.9% vs 29.8%, respectively; p = 0.0004), to be postpartum (66.7% vs 37.3%, respectively; p = 0.043), and to report treatment side effects (54.8% vs 30.1%, respectively; p < 0.0001). Reported treatment side effects (odds ratio [OR], 3.6; 95% confidence interval [CI], 2.2 to 6.2) and lack of medical insurance (OR, 1.7; 95% CI, 1.1 to 2.7) were each associated with treatment noncompletion in a model including both. Also, pregnant women were more likely than nonpregnant women to fail to initiate INH treatment (52.1% vs 14.7%, respectively; p < 0.0001). CONCLUSIONS: LTBI patients who are young, pregnant or postpartum, uninsured, and/or report treatment side effects may require additional case management to improve INH treatment completion rates.
BACKGROUND: The treatment of latent tuberculosis infection (LTBI) is essential for tuberculosis elimination in the United States, but the major limitation is poor adherence to therapy. To aid the design of targeted adherence interventions, we investigated the factors associated with noncompletion of isoniazid (INH) therapy for LTBI. METHODS: A retrospective analysis of patients with who failed to complete vs those who completed 9 months of INH therapy at the RISE TB Clinic (Miriam Hospital; Providence, RI) in 2003 was performed. Factors associated with treatment noncompletion were examined using univariate and multiple logistic regression analysis. RESULTS: Of 845 patients with LTBI, 690 patients (81.6%) initiated INH therapy, of whom 426 patients (61.7%) completed therapy, and 246 patients (35.6%) were lost to follow-up. Treatment was discontinued in 18 patients (2.6%). Patients who failed to complete therapy were younger (mean age, 30.6 vs 33.8 years, respectively; p = 0.006), and were more likely to be uninsured (42.9% vs 29.8%, respectively; p = 0.0004), to be postpartum (66.7% vs 37.3%, respectively; p = 0.043), and to report treatment side effects (54.8% vs 30.1%, respectively; p < 0.0001). Reported treatment side effects (odds ratio [OR], 3.6; 95% confidence interval [CI], 2.2 to 6.2) and lack of medical insurance (OR, 1.7; 95% CI, 1.1 to 2.7) were each associated with treatment noncompletion in a model including both. Also, pregnant women were more likely than nonpregnant women to fail to initiate INH treatment (52.1% vs 14.7%, respectively; p < 0.0001). CONCLUSIONS: LTBI patients who are young, pregnant or postpartum, uninsured, and/or report treatment side effects may require additional case management to improve INH treatment completion rates.