Risk Stratification and Factors Associated with Abandonment of Tuberculosis Treatment in a Secondary Referral Unit

PURPOSE: To study the factors associated with the risk of discontinuing active tuberculosis treatment among patients in an outpatient referral unit and to analyze the association between patients’ abandonment risk score and their odds of discontinuing the treatment. PATIENTS AND METHODS: In this cohort study, tuberculosis patients were prospectively followed up from June 2012 through July 2019 at a secondary tuberculosis referral unit in Mato Grosso do Sul, Brazil. At initial consultation, patients were interviewed using a standardized questionnaire and were assigned a score for the risk of treatment abandonment by the nurse. Univariate and multivariate analyses were performed using logistic regression. RESULTS: One hundred and forty-eight patients were included in the study, of which 65.0% (96/148) were male. Their mean age was 43.3 ± 14.8 years (range: 18–89 years). Smoking, drug use, repeated admissions, and a high abandonment risk score were the variables associated with the highest risk of discontinuing the treatment. The rate of tuberculosis and human immunodeficiency virus coinfection was 37.2%. The overall rate of global treatment abandonment was 10.8% (95% confidence interval [CI]: 6.1–16.2). Upon stratification of patients that abandoned by the risk score, 22.9% (8/35) of the ones that abandoned had a high risk, 10.9% (6/55) had an intermediate risk, and 3.5% (2/58) had a low risk of treatment abandonment. In multivariate analysis, the factors associated with abandoning the treatment were smoking [adjusted odds ratio (aOR) = 4.91 (95% CI: 1.08, 22.32)] and undergoing retreatment (aOR) = 3.66 (95% CI: 1.04, 12 88). CONCLUSION: Smoking and undergoing retreatment were independent risk factors for tuberculosis treatment abandonment in this center. Risk stratification can help prioritize the strengthening of treatment adherence among patients at higher risk of abandoning treatment in referral units.