INTRODUCTION: Although the aging population had been increasing in many countries, the factors associated with sputum conversion in elderly patients with pulmonary tuberculosis have not been fully elucidated. OBJECTIVES: We aimed to identify the predictors of delayed sputum conversion and to assess the impact of non-conversion on mortality during tuberculosis treatment in elderly patients. METHODS: Elderly patients (>65 years) admitted at our hospital in Japan for sputum smear-positive pulmonary tuberculosis were included. The risk factors for sputum non-conversion after 2 months of treatment were determined using multiple logistic regression. Cox hazard regression was used to assess the influence of non-conversion on mortality. RESULTS: We included 185 patients, with median age of 82 years (IQR, 79-88 years). The median time to conversion was 47 (95% CI 43-51) days, and 62 (34%) were identified as non-converters. Multivariate analysis showed that high pretreatment smear grade, high C-reactive protein level and poor performance status were associated with non-conversion. Non-conversion did not contribute to death during treatment. CONCLUSIONS: In elderly patients, inflammation level and physical activity level, along with initial smear grade may have a significant impact on delayed sputum conversion. Non-conversion after two months of treatment might not be related with mortality.
INTRODUCTION: Although the aging population had been increasing in many countries, the factors associated with sputum conversion in elderly patients with pulmonary tuberculosis have not been fully elucidated. OBJECTIVES: We aimed to identify the predictors of delayed sputum conversion and to assess the impact of non-conversion on mortality during tuberculosis treatment in elderly patients. METHODS: Elderly patients (>65 years) admitted at our hospital in Japan for sputum smear-positive pulmonary tuberculosis were included. The risk factors for sputum non-conversion after 2 months of treatment were determined using multiple logistic regression. Cox hazard regression was used to assess the influence of non-conversion on mortality. RESULTS: We included 185 patients, with median age of 82 years (IQR, 79-88 years). The median time to conversion was 47 (95% CI 43-51) days, and 62 (34%) were identified as non-converters. Multivariate analysis showed that high pretreatment smear grade, high C-reactive protein level and poor performance status were associated with non-conversion. Non-conversion did not contribute to death during treatment. CONCLUSIONS: In elderly patients, inflammation level and physical activity level, along with initial smear grade may have a significant impact on delayed sputum conversion. Non-conversion after two months of treatment might not be related with mortality.