Abstract: Objective  To explore the treatment dynamics and risk factors on treatment outcomes of pulmonary tuberculosis. Methods  Seven hundred and fifty-seven cases of primary pulmonary tuberculosis registered and followed up in Xiangshui county from 2009 to 2011 was analyzed by univariate and multivariate based on survival analysis model. Results  Of 757 patients, 720 cases were successfully treated. Median healing time of pulmonary tuberculosis was 182 days estimated by Kaplan-Meier survival curves. Univariate analysis on risk factor showed that the proportion of the observed numbers under successful treatment from detection were 24.17% (174/720)，66.53% (479/720)，9.31% (67/720) respectively. Compared to the proportion of the expected numbers under successful treatment (29.86%, 59.72%, 10.42%), the distribution difference was statistically significant (χ²=23.19, P=0.000). The proportion of the observed numbers under successful treatment from delayed diagnosis status were 26.67% (192/720) and 73.33% (528/720). Compared to the proportion of the expected numbers under successful treatment(31.25%，68.75%), the distribution difference was statistically significant (χ²=11.33, P=0.001). Cox multivariate analysis showed that factors with age under 30，（45－＜60，OR=1.213, 95%CI=1.014-1.452; ≥60（OR=1.323, 95%CI=1.069-1.636), being farmers(OR=0.787, 95%CI=0.650-0.953) and others (OR=0.755, 95%CI=0.591-0.966), with case detection by referring way(OR=1.498，95%CI=1.275-1.760), delayed diagnosis(OR=0.782，95%CI=0.658-0.929) affected the cure rate of pulmonary tuberculosis.  Conclusion  More attention should be paid to TB cases being farmers, with age under 30, detected by referring way and having delayed diagnosis.

Key words: Tuberculosis,pulmonary/therapy, Treatment outcome, Survival analysis